CURRICULUM VITAE. Danielle M. Cusmano. Degree and date to be conferred: Doctor of Philosophy, Drew University

Transcription

1 CURRICULUM VITAE Name: Danielle M. Cusmano Degree and date to be conferred: Doctor of Philosophy, 2014 Collegiate Institutions Attended: Drew University Bachelor of Arts, University of Maryland, Baltimore Doctor of Philosophy, 2014 Major: Neuroscience Peer-reviewed Publications: Cusmano, D.M., Hadjimarkou, M.M., & Mong, J.A. (2014) Gonadal steroid modulation of sleep and wakefulness in male and female rats is sexually differentiated and neonatally organized by steroid exposure. Endocrinology. 155: Cusmano, D.M. & Mong, J.A. (2014) In utero exposure to valproic acid changes sleep in juvenile rats: a model for sleep disturbances in autism. SLEEP. Accepted 03/07/2014. Professional Society Memberships: present Society for Neuroscience (SFN) present Society for Behavioral Neuroendocrinology (SBN) present Sleep Research Society (SRS) Awards and Honors: David and Pauline Bodle Endowed Scholarship Drew University Recognition Award Dean s List awarded for GPA of 3.4 and higher for eight semesters Schering Plough Endowed Scholarship 2008 Novartis Award in Neuroscience

2 2008 Honors in Neuroscience, summa cum laude, Drew University 2012 SRS First Time Attendee Travel Award 2012 SRS Abstract Meritorious Award Honorable Mention 2013 SRS Abstract Meritorious Award 2013 Greater Baltimore Chapter SFN Poster Award - First Place 2014 SRS Abstract Meritorious Award Honorable Mention 2014 Society for Women s Health Research Donald G. and Darel Stein Travel Award Research Support: 09/12-present F31AG043329, NIH/NIA, Modulation of REM sleep circuitry and involvement of hypocretin in female rats Cusmano, D.M. (PI)

3 Abstract Title of Dissertation: Identification of Key Preoptic Area Nuclei Mediating Sex Differences and Estrogenic Modulation of Sleep in a Rodent Model Danielle M. Cusmano, Doctor of Philosophy, 2014 Dissertation Directed by: Dr. Jessica A. Mong, Associate Professor, Department of Pharmacology Women are twice as likely as men to experience sleep disruptions and insomnia throughout their lifespan. The reason for this sex difference in sleep is unknown. In females, findings from clinical and basic research studies strongly implicate a role for gonadal steroids in sleep modulation. Sleep disruptions and complaints in women typically coincide with marked changes in the gonadal steroidal profile during puberty, the menstrual cycle, pregnancy, and menopause. Likewise, fluctuations in the ovarian hormonal milieu across the estrous cycle in female rodents correlate with changes in sleep and wakefulness, such that when estradiol levels are elevated sleep is suppressed. In male rodents, sleep remains relatively unchanged following changes in testosterone levels. Here, we sought to address the mechanism by which gonadal steroids selectively suppress sleep in females. Using a rat model, we found that sex differences in sleep resulted from activational effects of estradiol on sexually differentiated brain circuitry. Sexual differentiation of the brain organized the sleep circuitry, specifically the ventrolateral peroptic area (VLPO) and the median preoptic nucleus (MnPN). We found that estrogen receptor alpha expression in the MnPN was higher in females compared to males, making this nucleus an ideal target for estradiol s actions on sleep. In females,

4 antagonism of estrogen receptors within the MnPN attenuated the estradiol-mediated suppression of sleep. Furthermore, direct infusion of estradiol into the MnPN induced wakefulness and suppressed sleep. The orexinergic neurons in the lateral hypothalamus are regulated by MnPN activity and orexin is a key neuropeptide involved in arousal and suppression of sleep. We hypothesized that estradiol action in the MnPN would release the inhibitory tone on orexin but antagonism of orexin receptors did not completely block estradiol s suppressive effects on sleep. We also found a sex difference in the sleeppromoting effect of the dual orexin receptor antagonists, such that females experienced prolonged wake suppression compared to males. Together, these data begin to address major gaps in our knowledge regarding sex differences and hormonal modulation of sleep. Advancing our understanding of the mechanisms underlying hormonal modulation of sleep is imperative for better treatment of sleep disruption in women.

5 Identification of Key Preoptic Area Nuclei Mediating Sex Differences and Estrogenic Modulation of Sleep in a Rodent Model by Danielle M. Cusmano Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, Baltimore in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2014

6 Copyright 2014 by Danielle M. Cusmano All rights reserved

7 Acknowledgements I would like to thank my family and friends for all of their love and support over the years. To my parents, I cannot thank you both enough for all that you have done for me. To Jeremy, thank you for your love, encouragement, and understanding - I could not have done this without you! I would also like to thank the members of the Mong lab, past and present: Tamara, Michael, Mike, Mary, Joe, Shaun, Sarah, Katrina, and Jenny. I am so grateful to have worked with each of you. To my committee: Dr. Margaret McCarthy, Dr. Istvan Merchenthaler, Dr. Asaf Keller, and Dr. Gloria Hoffman. Thank you all for the support and guidance throughout this project and for pushing me to always do my best. To Dr. Jessica Mong, thank you for everything! I could not have asked for a better mentor. You have been a wonderful teacher and friend. I have learned so much from you and would not be the scientist I am today without you. iii

8 TABLE OF CONTENTS ACKNOWLEDGEMENTS... III LIST OF TABLES... VI LIST OF FIGURES... VII ABBREVIATIONS... X CHAPTER 1: GENERAL INTRODUCTION... 1 SLEEP... 1 CHARACTERISTICS OF SLEEP... 3 SLEEP- WAKE CIRCUITRY SEX DIFFERENCES IN CIRCADIAN RHYTHMS AND SLEEP HORMONAL INFLUENCES ON SLEEP ACROSS THE LIFESPAN RODENT MODELS OF SLEEP INFLUENCES OF SEX AND HORMONES ON THE SLEEP CHAPTER 2: GOALS OF THE DISSERTATION CHAPTER 3: GENERAL METHODS ANIMALS SURGERIES HORMONE REPLACEMENT PARADIGM DATA ACQUISITION IMMUNOCYTOCHEMISTRY CHAPTER 4: NEONATAL EXPOSURE TO GONADAL STEROIDS ORGANIZES ADULT SLEEP BEHAVIOR INTRODUCTION METHODS RESULTS DISCUSSION CHAPTER 5: THE SLEEP-ACTIVE PREOPTIC AREA AS A KEY SITE IN ESTRADIOL S SUPPRESSION OF SLEEP INTRODUCTION iv

9 METHODS RESULTS DISCUSSION CHAPTER 6: OREXIN IS NOT THE MEDIATOR OF ESTRADIOL S EFFECTS ON SLEEP INTRODUCTION METHODS RESULTS DISCUSSION CHAPTER 7: IN UTERO EXPOSURE TO VALPROIC ACID CHANGES SLEEP IN JUVENILE RATS: A MODEL FOR SLEEP DISTURBANCES IN AUTISM INTRODUCTION METHODS RESULTS DISCUSSION CHAPTER 8: GENERAL DISCUSSION SUMMARY USING A RODENT MODEL FOR STUDYING ESTRADIOL S EFFECTS ON SLEEP DEVELOPMENTAL ORGANIZATION OF SLEEP IMPACT OF SEX AND HORMONAL STATUS ON SLEEP LOSS THE MNPN AS A KEY SITE OF ESTRADIOL S EFFECTS ON SLEEP THE MNPN AS A POTENTIAL SITE MEDIATING SLEEP DISRUPTIONS IN MENOPAUSE OTHER POTENTIAL SITES FOR ESTRADIOL S SUPPRESSION OF SLEEP IMPLICATIONS FOR HYPNOTIC USE AND DEVELOPMENT CONCLUSION REFERENCES: v

10 List of Tables Table 1. Effects of sex-specific gonadal steroid replacement on sleep-wake behaviors during the light phase Table 2. Effects of DHT-replacement on the duration of sleep-wake behaviors in females and males during the dark phase Table 3. Effects of direct infusion of ICI into the MnPN on the duration of sleepwake behaviors in oil and estradiol-treated females Table 4. Effects of direct infusion of ICI into the VLPO on sleep-wake behaviors in oil and estradiol-treated females Table 5. Effects of in utero exposure to VPA on vigilance state transitions Table 6. Effects of in utero exposure to VPA on averaged power spectra (% total power) in the light and dark phase Table 7. Effects of in utero exposure to VPA on mean power (% total power) for alpha, sigma, and beta frequency bands vi

11 List of Figures Figure 1. Representative traces of EEG and EMG waveforms from a telemetric recording... 4 Figure 2. Sleep characteristics... 7 Figure 3. Sleep-wake circuitry Figure 4. Fluctuation in ovarian hormonal milieu across the menstrual cycle Figure 5. Representative hypnograms of rodent sleep-wake behavior Figure 6. Fluctuation in ovarian hormonal milieu across the estrous cycle Figure 7. Organizational-Activational hypothesis of hormone action Figure 8. Estradiol has genomic and non-genomic mechanisms of action Figure 9. Gonadal steroid replacement paradigm Figure 10. Sleep-wake behavior in GDX female and male rats Figure 11. Sleep-wake behavior following sex-specific gonadal steroid replacement in female and male rats Figure 12. Effects of sex-specific gonadal steroid replacement on sleep-wake behavior in female and male rats Figure 13. Magnitude of change in vigilance states following gonadal steroid replacement Figure 14. Effects of estradiol and testosterone replacement on sleep-wake behavior in females, males and masculinized females Figure 15. Magnitude of change following estradiol and testosterone on sleep and wake in females, males and masculinized females Figure 16. Developmental organization of Fos expression in the VLPO vii

12 Figure 17. ER alpha expression in the MnPN and VLPO Figure 18. Injection and infusion paradigm for ER-antagonism studies Figure 19. Effects of direct infusion of ICI into the MnPN on estradiol-mediated changes in sleep Figure 20. Effects of direct infusion of estradiol into the MnPN on sleep and wakefulness Figure 21. Effects of direct infusion of estradiol into the MnPN on sleep and wakefulness across the light and dark phase Figure 22. Effects of direct infusion of ICI into the VLPO on estradiol-mediated changes in sleep and wake Figure 23. Effects of a systemic injection of ICI on sleep and wake following estradiol Figure 24. Gavage and hormone replacement paradigm Figure 25. DORA-12 dose response Figure 26. Effects of DORA-12 in oil and estradiol-treated females Figure 27. Effects of DORA-12 across the 12h dark period Figure 28. The sleep-promoting effect of DORA-12 is sex dependent Figure 29. Effects of hormonal status on DORA-12 efficacy Figure 30. Effects of ICI infusion in the MnPN on orexinergic activation Figure 31. Representative hypnograms of juvenile rats exposed to SAL and VPA in utero across a 24h period Figure 32. Effects of in utero exposure to VPA on sleep-wake behavior in juvenile rats viii

13 Figure 33. Effects of in utero exposure to VPA on sleep-wake architecture in juvenile rats Figure 34. Effects of in utero exposure to VPA on cortical power spectra in juvenile rats Figure 35. Effects of in utero exposure to VPA on cortical expression of GAD65 and GAD67 in juvenile rats ix

14 Abbreviations ARAS ASD BF DRN EB EEG ER GABA GDX ascending reticular activating system Autism Spectrum Disorders basal forebrain dorsal raphe nucleus estradiol benzoate electroencephalogram estrogen receptor gamma-aminobutyric acid gonadectomy ICI ICI 180,270 KPBS L-PGDS LC MCH MnPN NREMS ORDX OVX PF/LH POA PSG REMS potassium phosphate buffered saline lipocalin-type prostaglandin D2 synthase locus coeruleus melanin concentrating hormone median preoptic nucleus non-rapid eye movement sleep orchidectomy ovariectomy perifornical area/ lateral hypothalamus preoptic area polysomnography rapid eye movement sleep x

15 SAL SWA SWS TBS TMN vlpag VLPO VPA ZT saline slow wave activity slow wave sleep tris-buffered saline tubberomammilary nucleus of the hypothalamus ventolateral periaqueductal grey ventolateral preoptic area valproic acid zeitgeber xi

16 Chapter 1: General Introduction For decades, scientists have been plagued with questions of how and why we sleep. Until the early 20 th century, sleep was thought to be a passive process, caused by the cessation of stimuli that maintain consciousness. The first sleep-active nucleus was discovered less than 20 years ago, and since its discovery our understanding of how sleep is initiated and regulated has grown at a rapid pace. However, many unanswered questions still remain regarding underlying mechanisms of sleep and even its function. Much of what we know about sleep has been generalized to the male physiology. Women and female lab animals have been underrepresented in sleep research, even though sleep complaints are twice as prevalent in women throughout all stages of life. The cause of sex differences in sleep remains unknown. Sleep Sleep is defined as a reversible physiological state of reduced consciousness and activity. Behavioral states defined as sleep or sleep-like should meet the following criteria: (i) a state of inactivity that reoccurs on a daily basis, (ii) a stable but reversible state that has an increased arousal threshold to stimuli, (iii) a state that is regulated by homeostatic responses, (iv) a state specified stereotypic posture that is species-specific, and (vii) state-related changes in neural functioning [for review (1,2)]. Based on these criteria, all organisms in the animal kingdom, from insects, mollusks, fish, reptiles, amphibians, birds, to mammals experience daily states of sleep-like behavior [for review (1,3)]. 1

17 The amount and timing of sleep varies greatly between species but the presence and need for sleep is evolutionarily conserved [for review (1)]. Incredibly diverse sleep patterns exist even within mammals that have the more classically defined sleep periods [for review (1)]. Diurnality influences when in the 24h day sleep occurs. Diurnal animals sleep during the night, whereas nocturnal animals sleep during the day and are active at night. Sleep can also be consolidated (monophasic) or fragmented (polyphasic). Humans have consolidated periods throughout the night that average ~7-8 hours. Rodents have fragmented sleep, cycling though sleep and wake quickly. Rats sleep about hours per day but their average sleep bout duration is ~3-5 minutes. They accumulate the majority of their sleep during the quiescent phase, but still sleep during their active phase. It is unclear why this diversity exists. One hypothesis is predation pressure (4). The reduced state of consciousness characteristic of sleep leaves organisms vulnerable to their environment. Animals that are easily preyed upon such as rodents must frequently survey their environment and thus have adapted sleep patterns that are interrupted by shorts period of wakefulness. Marine mammals have a unique form of sleep that occurs only in one hemisphere of the brain. This type of sleep may have evolved to allow for surfacing to breathe while sleeping [for review (5)]. Daily reoccurrence of sleep is controlled by endogenous circadian, or daily, rhythms [for review (6)]. Circadian rhythms are endogenous changes that occur in biological processes throughout a period of about 24h. The suprachiasmatic nucleus (SCN) is the brain s master clock and pacemaker and controls circadian rhythms [for review (7)]. Sleep is just one type of circadian rhythm. There are also rhythms in core body temperature and melatonin. Body temperature is elevated during the day and drops 2

18 at night. Melatonin, the sleep hormone, is produced by the pineal gland a couple hours before an individual s bedtime and peaks around 2-3AM. These rhythms can occur in the absence of external cues because the SCN has an endogenous rhythm. For example, periods of sleep and wakefulness will occur even in the absence of light. Rhythms can be entrained, or reset, around a given stimulus, like light or food [for review (8)]. Despite circadian influences on sleep, the SCN does not have direct connections to the sleep circuitry. The duration of sleep remains unchanged in animals with SCN lesions, but the sleep is fragmented (9-11). Although the circadian system influences sleep timing, sleep behavior is mediated by sleep-specific factors and circuitry. Characteristics of Sleep Before advances in technology, sleep was thought to be due to termination of brain activity and ultimately a state of unconsciousness (12). With the advent of electroencephalography (EEG), it became clear that the brain exhibits characteristic activity during sleep (13). The EEG consists of multiple surface electrodes that measure brain activity. The output is the summation of synaptic activity of the cortical neurons below the surface of the skull. Using this device, Hans Berger discovered that the brain was not quiet during sleep. He described the wake EEG recording as desynchronized, being composed of two types of waveforms: alpha and beta waves (Fig. 1) (13). These waveforms are high frequency (8-12Hz and 16-24Hz, respectively), low amplitude waves. Alpha waves typically occur during quiet wake and periods of relaxation and drowsiness, while beta waves are more prominent during active wake and consciousness. The transition from wake to sleep is often characterized by the changes in the alpha 3

19 rhythm. Unlike the wake EEG, the sleep EEG displays slow, synchronized cortical activity, not the absence of activity as was predicted (Fig. 1). We now know that cortical activity is driven by thalamocortical system and forebrain projections so EEGs reflect the underlying activity of these brain regions. Figure 1. Representative traces of EEG and EMG waveforms from a telemetric recording. Sleep Stages Loomis, Harvey, and Hobart were the first to describe sleep based on characteristic EEG activity (14). They noticed 5 distinct stages of EEG activity: alpha, low voltage, spindles, spindles plus random, and random. These brain waves characterize the sleep state known as non-rapid eye movement sleep (NREMS). NREMS consists of 3 stages (N1-N3), ranging from light sleep (N1) to deep sleep (N3) (15). Stage N1 is a relaxed state of drowsiness; individuals have slow eye movements and can be easily aroused from stage N1 sleep. The stage N1 EEG is composed predominately of theta waves in the 4-8Hz frequency range. Stage N2 is a deeper sleep than N1, but individuals are still easily awakened. During stage N2, the EEG is characterized by the presence of K-complexes and sleep spindles. K-complexes are high amplitude bi- or tri-phasic waves and can be both spontaneous and evoked by sensory stimuli (16). K-complexes are common during the transition into stage N3 and proposed to protect the sleep state from 4

20 sensory stimuli [for review (17)]. Sleep spindles are quick bursts of activity in the 11-16Hz frequency range generated in the thalamus (18). Sleep spindles are thought to be involved in synaptic plasticity and learning and memory and, like K-complexes, are believed to protect the sleep state from sensory stimuli [for review (19)]. Stage N3, also called slow wave sleep (SWS), is the deepest stage of sleep. It is characterized by the predominance of the low frequency and high amplitude delta waves in the EEG. Quantification of these slow waves is referred to as slow wave activity (SWA) and considered a neurophysiological marker of sleep depth and intensity. A second sleep state was discovered serendipitously by Aserinsky and Kleitman in the early 1950s (20). Researchers did not typically stay for observation after their subjects began to drift to sleep. Aserinsky was monitoring recordings from his son when he noticed transitions in his sleep EEG and rapid, jerky eye movements not reminiscence of sleep (21). This was the first observation of rapid eye movement sleep (REMS). Further investigation of the sleep state revealed that it is also associated with increased cardiac and respiratory rates as well as dreaming (20). In 1959, work by Jouvet and colleagues in animals demonstrated that REMS is unique from NREMS, as it is characterized by high frequency EEG and muscle atonia (Fig. 1) (22). The EEG during REMS is more reminiscent of wakefulness and was thus termed Le sommeil paradoxal, paradoxical sleep, by Jouvet (22). More recently, EEG has been combined with other physiological measurements that include muscle tone, eye movements, heart rate and respiratory rate, for a more complete measure of sleep behavior. Polysomnography (PSG) incorporates these physiological parameters to determine stage, depth, intensity, and quality of sleep. 5

21 Muscle tone is reduced from wake to NREMS and absent during REMS. Rapid eye movements are characteristic of the REMS state. Additionally, both cardiac and respiratory rates are lower during NREMS than wake and REMS. These physiological measures can also aid in the diagnosis of sleep disorders. For example, sleep apnea is a disorder where sleep is disrupted by pauses in breathing. A PSG sleep study can identify the duration, timing and sleep-stage association of these events. Sleep Patterning Sleep occurs in cycles throughout the night. Individuals typically transition through the stages of NREMS from light sleep to deep sleep (N1-N3) and then to REMS. On average, sleep cycles occur about every 90 minutes, roughly 3-5 times per night (23). Stage N3 is more prominent during the early part of the night, while REMS periods increase in duration as the night progresses (Fig. 2A). The amount of time spent in each sleep phase is strongly influenced by age (Fig. 2B) (24,25). Following birth, infants spend more than 14 hours asleep. About 50% of that time is spent in REMS. The amount of REMS sharply declines following early development and remains relatively stable until older age (24,26). During the adolescent period, sleep and EEG sleep characteristics undergo significant changes. Studies consistently find delays in sleep onset, or shifts to later bedtimes during adolescence [for review (27,28)]. Total sleep is reduced during this time (29-31), but it is mainly due to a significant decrease in SWS by about 40% (32). Delta power is also reduced (33-35), and this reduction occurs earlier in girls than in boys, mainly due to the earlier onset of puberty in girls (33,36). This decline in SWS is associated with brain reorganization and 6

22 Figure 2. Sleep characteristics. (A) Individuals cycle through the stages of sleep multiple times throughout the night. NREMS stage N3, or SWS, is longer and occurs earlier in the night while REMS increases in duration and frequency as the night progresses. (B) Sleep declines with increasing age. Infants have the highest amount of total sleep and REMS. There is a sharp decline in REMS following infancy, which remains relatively stable before it begins to slowly decline later in life. Total sleep time (TST), on the other hand, continually declines with increasing age. Figure was modified from Rofffwarg et al [1966] Science. 7

23 maturation during adolescence. The decrease in delta is hypothesized to be due to reductions in synaptic density and cortical grey matter (36-38) [for review (39)]. It is unclear if hormonal factors during puberty significantly influence the cortical changes underlying the reduction in delta power. In the elderly population, the intensity and quality of sleep is greatly reduced (24,25). Older individuals commonly report disrupted sleep, which is characterized by increased sleep latency, frequent night awakenings, and daytime sleepiness. Stage N1 and N2 sleep are increased and SWS is decreased (40). Due to frequent sleep difficulties, the use of sleep aids is also more common in the aged population (41,42). These changes in sleep may be due to age-related changes in sleep-wake circuitry activity or circadian rhythms, decreases in homeostatic sleep pressure, or comorbidities like depression [for review (43)]. Function of Sleep Sleep has been extensively studied, yet the function of sleep remains elusive. One prevailing theory is that sleep serves a restorative function for the brain and body. Energy expenditure is high during wake. Neuronal activity requires energy, specifically ATP. It is metabolized to adenosine, accumulates with wakefulness and dissipates during sleep. The accumulation of adenosine is implicated in the onset of sleep [for review (44)]. Sleep may serve as a time to restore the cellular pools of energy by reducing energy utilization and by facilitating the recycling of adenosine back to ATP [for review (45)]. The synaptic homeostasis hypothesis proposes that sleep is a period of renormalization [for review (46)]. Synapses are pruned and strengthened to form stable memories and cellular 8

24 homeostasis is restored [for review (46,47)]. Many studies have highlighted the importance of sleep in brain plasticity associated with learning and memory. Sleep is critical for the formation and consolidation of new memories [for review (48,49)]. Memory retention and performance of learned tasks improves after periods of sleep (50-55). Electrophysiological correlates of learning and memory, particularly long term potentiation (LTP) in hippocampal neurons, are attenuated by periods of sleep deprivation (56-59), specifically cyclic adenosine monophosphate (camp) and protein kinase A (PKA)-dependent LTP (60). At a molecular level, sleep loss impairs camp- PKA signaling by increasing the amount and activity of enzyme responsible for camp degradation, phosphodiesterase (PDE4), and ultimately leading to decreased phosphorylation of the downstream target, camp response element binding protein (CREB) (60). These findings implicate that the cellular and molecular underpinnings of memory and memory formation are regulated by sleep. A recent study emphasizes a critical role for sleep in cellular homeostasis. During sleep, the interstitial space of the brain increases by more than 60% enabling clearance of metabolic byproducts by the glymphatic, or brain lymphatic, system in rats (61). Xie and colleagues report that sleep improves the clearance of amyloid beta, which is a toxic cellular byproduct and component of protein aggregates associated with Alzheimer s disease. Waste clearance may play a key part in the restorative function of sleep and may aid in the prevention of neurotoxicity associated with neurodegeneration. The sleep theories described above mainly focus on a functional role for NREMS, particularly SWS. It is unclear if REMS significantly contributes to these functional roles of sleep. Two main functional hypotheses for the role of REMS include the ontogenetic 9

25 and readying the brain for wake hypotheses. REMS is high during early development and drastically declines with age, implicating a role for REMS in brain development and the formation of mature circuits which is readily occurring during this time (24). Roffwarg and colleagues hypothesized that the brain stimulation during REMS assisted in developmental processes (24). It is now understood that REMS enhances cortical plasticity during development which enables the sculpting of the sensory systems in the brain [for review (62)]. But REMS occurs in adults as well so what role does it play during adulthood? Studies investigating the effects of sleep loss on neurogenesis suggest that REMS may be important hippocampal cell proliferation [for review (63)]. The prevailing REMS theory is that REMS prepares the brain for wakefulness [for review (3)]. REMS increases towards the end of the night and individuals typically awaken from a REM episode. Therefore, it is hypothesized that the REMS state aids the brain in recovering from sleep and helps get it back online. REMS deprivation due to lesions or monoamine oxidase inhibitors does not lead to any difficulties waking up, challenging this longstanding hypothesis for the role of REMS. Consequences of Sleep Loss Studies investigating the consequences of sleep loss highlight the importance of sleep for both the brain and body. Partial sleep deprivation can result from sleep fragmentation (frequent night awakenings disrupting sleep), selective sleep state loss (NREMS vs. REMS) or sleep restriction (reduced sleep duration) [for review (64)]. Sleep restriction commonly occurs in our society. It causes impaired alertness and cognitive functioning, and impaired physiological responses including cardiovascular, metabolic, 10

26 immunological functioning [for review (64,65)]. Chronic insufficient sleep also increases one s risk for affective mood disorders like depression and anxiety [for review (66,67)], and mortality (68,69). Sleep loss in children has similar effects on physiology and cognition. Childhood insufficient sleep is associated with behavioral problems (70,71) [for review (72)], anxiety and depression (70,73), as well as metabolic problems associated with an increased risk for obesity (74). Studies using animal models indicate that sleep loss during development impairs sensory system development and can affect adult behaviors. In fruit flies, learning (75) and courtship behaviors (76) are impaired due to sleep loss during developmental and in rats, selective REMS loss reduces adult male sexual behavior (77). There is a prominent role for sleep in endocrine functioning so it is not surprising that sleep loss has profound effects on endocrine regulated processes. Many hormones are secreted during sleep. Growth hormone, for example, is secreted during sleep and is predicted to involved in the restorative function of sleep (78). Conversely, cortisol, the stress hormone, is reduced during sleep and rises just before the onset of wake. Metabolic function is strongly influenced by sleep [for review (79)]. Sleep loss disrupts satiety signals and alters glucose metabolism and tolerance. This increases the risk for obesity, insulin resistance and type II diabetes. [for review (80)]. Additionally, sleep influences the activity of the hypothalamic-pituitary-gonadal axis, which is important for the normal reproductive functioning. Shifts in sleep patterns like during jetlag or shift work are associated with menstrual disturbances and reproductive dysfunction [for review (81)]. 11

27 These are just a few examples of how chronic insufficient can have adverse effects endocrine system function. Sleep-Wake Circuitry It was long assumed that sleep was caused by the lack of stimuli or signaling to the areas that control wakefulness, and therefore, a passive process. In 1930, Constantin von Economo proposed that sleep was actually an active process, regulated by distinct brain areas (12). During the late 1910 and early 1920s, there was an outbreak of the encephalitis lethargica. The cause was unknown, but now possibly linked to a rare streptococcal infection that induces a strong immune response (82). von Economo observed patients that showed behaviorally distinct symptomology. One group experienced extreme sleepiness, while another had insomnia. Lesions were found in patients with both types of sleep disruptions, localized anteriorly in the lateral wall of the third ventricle, near the corpus striatum [and] in the posterior wall of the third ventricle near the nuclei of the oculomotorius in the cap of the interbrain for patients with insomnia and extreme sleepiness, respectively (12). Based on his observations, he states: we must insist on the anatomical fact that the center for regulation of sleep is in the immediate vicinity of the other important vegetative centers located in the infundibular region [ ] (12). It was not until 66 years after von Economo s proposal that the sleep-switch was identified in the preoptic area. Sleep-Switch Key to the existence of a sleep-switch is the identification of a nucleus or nuclei that is/are (i) active during sleep but not wake, (ii) innervate and inhibit wake-promoting 12

28 nuclei, and (iii) when inhibited, will induce arousal. In 1996, Sherin and colleagues identified the ventrolateral preoptic area (VLPO) in rats as a sleep-active nucleus and potentially the sleep-switch von Economo proposed (83). This was the first anatomical characterization of a sleep-active area. VLPO neurons express the Fos protein, a proxy marker for recently activated neurons (84), during sleep (83,85-88) and show increased firing rates during both NREMS and REMS sleep compared to wake (89). These neurons are not active during periods of increased sleep propensity, or sleep need, like during sleep deprivation, suggesting that the VLPO is not involved in sleep homeostasis (83,89). Lesions of the VLPO reduce total sleep time (90). NREMS loss is linearly correlated with the number of cells lost within the VLPO cluster. Near complete ablation of the VLPO cluster (~80-90 cell loss) reduces NREMS by ~50% and a near 60% reduction in EEG delta power (90). Lesions to the extended portion of the VLPO (evlpo), which is dorsal and medial to the VLPO cluster, are not associated with reductions in NREMS but rather to reductions in REMS, indicating that the role of VLPO neurons in sleep regulation is different for subpopulations of VLPO cell groups (90). This notion is further supported by more recent work indicating that a subpopulation of VLPO cells is responsive to changes in REMS homeostasis (91), since they increase wake firing during short-term sleep deprivation (92). The VLPO sleep-active neurons contain the inhibitory neurotransmitter gammaaminobutyric acid (GABA) and the inhibitory neuropeptide galanin (85,93). These inhibitory neurons strongly innervate brain regions associated with arousal (Fig. 3A) [for review (94,95)], including the tuberomammilary nucleus [TMN, (83,85,96)], the lateral hypothalamus [LH, (96-99)], and brainstem nuclei like the dorsal raphe nucleus [DRN, 13

29 (96,100)] and locus coeruleus [LC, (96)]. The VLPO receives reciprocal inhibitory projections from these arousal nuclei (101). It is this reciprocal, inhibitory connectivity that is postulated to mediate the switching between sleep and wake and referred to as the sleep-switch (95,102). The median preoptic nucleus (MnPN) is the only other identified sleep-active nucleus. Like the VLPO, the MnPN contains GABAergic sleep-active cells, as well as glutamatergic cells (103). The MnPN is a critical nucleus in physiological homeostasis, as it is involved in temperature, body fluid and cardiovascular regulation [for review (104,105)]. The MnPN is located adjacent to and connected with circumventricular organs (106) that lack a complete blood brain barrier and transport humeral factors to the brain. These circumventricular organs include the organum vasculosum of the lamina terminalis and the subfornical organ. This situates the MnPN in such a way as to respond to factors influencing homeostasis from the brain and body. The MnPN is a key site for sleep homeostasis. The firing rates of sleep-active MnPN neurons increase just prior to sleep, during NREMS and increase even further during REMS (107). Similar to the VLPO, Fos expression in the MnPN is also correlated to the preceding amount of sleep acquired, however; these neurons also express Fos during sleep deprivation (86-88,108). In fact, the number of Fos-positive GABAergic cells is greatest when sleep propensity is at its peak. The activity of these neurons indicates that the MnPN is sensitive to changes in sleep pressure making it a key regulator of sleep homeostasis. The MnPN innervates similar arousal nuclei as the VLPO including the LH, DR, and LC as well as the periaqueductal gray (vpag) and the magnocellular preoptic area 14

30 Figure 3. Sleep-wake circuitry. Sleep-active cells primarily found within the (A) VLPO and (B) MnPN project to and inhibit arousal nuclei to promote sleep. (C) REMS-ON in the brainstem are under inhibitory control of the REMS-OFF cells. When the REMS-OFF cells are inhibited and the inhibitory tone is release on the REMS-ON cells, they become activated and induce REMS. (D) Wake-promoting nuclei within the ascending reticular activating system of the brainstem project to the hypothalamus and forebrain where they join arousal projections like histaminergic and cholinergic neurons from the TMN and BF, respectively, to promote wake. These regions also actively inhibit sleep-promoting nuclei in the POA. Adapted from Brown et al. [2012] Physiological Reviews. 15

31 (Fig. 3B) (109). GABAergic MnPN neurons have direct inhibitory control over the orexinergic neurons in the perifornical area/ lateral hypothalamus (PF/LH) (110). These orexinergic neurons are a key source of arousal signaling, suggesting a sleep-promoting mechanism of the MnPN. The MnPN also densely innervates the VLPO and may influence sleep by regulating the activity of sleep-active VLPO neurons (101). Within the LH, a subpopulation of sleep-active cells was identified (111). These neurons express the cyclic neuropeptide melanin-concentrating hormone (MCH) and intermingled with neurons expressing the arousal neuropeptide orexin ( ). Intracerebroventricular administration of MCH induces a dose-dependent increase in both SWS and REMS, but the promotion of REMS is greater (114). Pharmacological antagonism of the MCH receptor induces wakefulness at the expense of SWS and REMS (115). Genetic knockout of the MCH receptor report, however, produced conflicting results. One study found that MCH receptor knockout mice have increased wakefulness and decreased sleep (116), while another reports excess sleepiness in knockout animals (117). The exact role of the MCH neurons in sleep is unclear. Recent optogenetic studies suggest that MCH plays a key role in sleep onset and intensity (118), involved in the transition from NREMS to REMS (119), and help maintain REMS (120). Stimulation of MCH cells reduces sleep onset, the duration of wake bouts, and increases EEG delta power (118). Stimulation during NREMS sleep induces transitions from NREMS to REMS (119) and activation of MCH cells or MCH terminals in the TMN increases the duration of the REMS bouts (120). Silencing of MCH cells does not affect vigilance state, but controlled ablation via cell-specific expression of diphtheria toxin A increases wake and suppresses NREMS only (119). 16

32 The circuit mediating REMS is localized to the brainstem (Fig. 3C). In the rat, three REMS-ON, or REMS-active, nuclei have been identified by their increased Fos expression during periods of REMS. These regions include the sublaterodorsal nucleus (SLD) and the precoeruleus region (PC) (121,122). Direct infusion of GABA receptor antagonists into the SLD disinhibits the SLD and its activation induces REMS (121). Lesions of the SLD reduce and fragment REMS (122). The REMS generating neurons of the SLD are glutamatergic and induce the muscle atonia associated with REMS (123,124). Descending SLD neurons project to premotor neurons in the ventral medulla, which inhibit motor neurons in the spinal cord, or directly to the spinal cord to produce atonia (122). The SLD also has GABAergic neurons that project to REM-OFF regions (122). REMS-OFF cells are localized in the ventrolateral periaqueductal gray (vlpag) and lateral pontine tegmentum (LPT) and have GABAergic projections to REMS-ON regions and exert inhibitory control over REMS-ON cell activity (121,122). Chemical inactivation and lesions of this region increase REMS (122,125,126). The forebrain sleep-switch, particularly the evlpo, sends inhibitory projections to the vlpag/ltp to potentially aid in the switch to REMS, while orexinergic neurons, which inhibit REMS, also project to and activate vlpag/ltp neurons (122). Cortical activation during REMS is generated by activation of the cholinergic neurons of the brainstem. The pedunculopontine and laterodorsal tegmental nuclei (PPT/LDT), Electrical stimulation of these neurons increases REMS (127,128) and lesions significantly reduce it (129,130). Two hypotheses have been proposed regarding REMS switching. The first is that there is a REMS-switch mediated by the reciprocal inhibitory connections between REMS-ON and REMS-OFF (122). This model has been proposed because the reciprocal 17

33 inhibitory connections would allow for the quick transitioning between NREMS and REMS. The second hypothesis is that MCH/GABAergic neurons in the LH initiate REMS by inhibiting the REMS-OFF cells in the vlpag [for review (131)]. A large number of Fos+ cells are found in the posterior hypothalamus, particularly in the PF/LH, following during REMS deprivation. About 75% of those cells were identified as being GABAergic and 1/3 of those also express MCH (114,132). It is currently unclear what triggers the switch into REMS. However, optogenetic activation of MCH neurons during NREMS triggers a transition into REMS, suggesting that the MCH/GABAergic neurons initiate the switch (119). Wake-promoting nuclei In the early 1930s, Frederic Bremer discovered where the nuclei involved in the maintenance of arousal were localized. He induced sleep-like behavior and delta waves following transection at the mid-collicular level of the midbrain in cats but transection at the mid-pons or lower did not affect wake (133) [for review (94,134)]. Moruzzi and Magoun supported these findings by inducing a wake-like EEG through stimulation of the reticular formation of the brainstem (135) [for review (94)]. The nuclei involved in the generation of arousal have been collectively called the ascending reticular activating system (ARAS). The ARAS is a network of brainstem nuclei involved in forebrain activation during wake and behavioral arousal (Fig. 3D) [for review (94)]. There are two pathways making up the ARAS. The first includes brainstem nuclei that induce cortical activation via a relay through the thalamus. The nuclei include the cholinergic PPT/LDT, which are active during wake and REMS, and exhibit little 18

34 activity during NREMS ( ). The EEG during wake and REMS is similar so it is not surprising then that the PPT/LDT are active during both behavioral states. The second ARAS pathway consists of brainstem nuclei that project to the hypothalamus and to the cortex. These nuclei include the noradrenergic neurons in the LC, the serotonergic neurons of the DRN and median raphe nucleus (MRN), and dopaminergic neurons in the vlpag. The noradrenergic neurons of the LC are most active during wake, decrease their firing during NREMS and are essentially silent doing REMS (139,140). Serotonergic neurons of the DRN and MRN exhibit similar firing patterns, such that they fire during wake, slow firing during NREMS, and cease firing during REMS (141,142). The dopaminergic neurons in the vlpag are primarily wake active and loss of these neurons significantly reduces wake by >20% (143). Lesions of neither the LC nor the DRN/MRN cause significant reductions in wake, suggesting that multiple pathways are involved in the induction and maintenance of arousal [for review (94,134)]. These brainstem projections along with hypothalamic and basal forebrain (BF) projections are involved in the onset and maintenance of arousal. In the hypothalamus, the histaminergic neurons of the TMN and the orexinergic neurons in the PF/LH promote arousal. The histaminergic neurons exclusively fire during wake (144,145). Pharmacological and genetic reduction of histamine signaling increases sleep, while enhancing signaling increases wake ( ). Orexinergic neurons in the PF/LH are also more active and release more orexin during wake and towards the end of REMS ( ). Orexin, or hypocretin, is neuropeptide involved in arousal. The loss of orexin leads to the sleep disorder narcolepsy, in which sleep and wakefulness are abnormally regulated [for review (153)]. Periods of wake are punctuated by extreme and 19

35 sudden sleepiness and transitions into REMS. Animal models of narcolepsy have revealed the key role orexin plays in vigilance state stabilization ( ). Orexin promotes consolidated wakefulness and also inhibits REMS (155,156, ). The basal forebrain contains a group of cholinergic neurons that are wake-active and involved in high frequency cortical activation. These neurons fire primarily during wake and REMS and are involved in cortical activation, especially in the theta (4-8Hz ) and gamma (> 30Hz) frequency ranges (163). Wake duration is unaffected by lesions of these neurons, but cholinergic loss reduces gamma frequency oscillations in the EEG (164,165). The basal forebrain also contains a population of GABAergic neurons that project to the cortex and inhibit cortical interneurons during wake, which leads to cortical activity (166). Clinical and animal models have been profoundly useful in elucidating the circuitry underlying sleep and wake. The main focus of current research is to understand the factors that influence sleep and affect the activity of the underlying circuitry. Many compounds, like adenosine, cytokines, neurotransmitters, and prostaglandins were found to influence sleep [for review (167)]. But one currently understudied area of research is on sex differences in sleep. We know so little about basic sex differences and gonadal steroid modulation of sleep. It is unclear what differences are present and how gonadal steroids play a role. The main goal of this dissertation is to begin to explore some of these unanswered questions, to ask how sex differences in sleep occur, if the sleep circuitry is different, and how and where gonadal steroids act to modulate sleep. 20

36 Sex Differences in Circadian Rhythms and Sleep Biological sex influences circadian rhythms in humans. The timing of the circadian rhythms for melatonin and core body temperature occurs earlier in women compared to men (168). The duration of the endogenous circadian period, or tau, is shorter in women compared to men (169) and the entrainment of the rhythms are different in men and women (168). The circadian timing of sleep is different as well. Women tend to go to bed earlier and wake up earlier than men (170,171). This difference is apparent during puberty and remains until menopause [for review (172)]. In addition to the timing of sleep, subjective and objective sleep parameters vary amongst the sexes. Typically, women complain of disrupted and insufficient sleep more frequently than men ( ). They report poorer sleep quality, difficulties falling asleep, frequent night awakenings and longer periods of time awake throughout the night (173,177). While subjective sleep studies suggest that women sleep worse than men (41,173,176,177), objective sleep studies indicate that women actually sleep better than men, having higher amounts of total sleep, more SWA, better sleep quality and a slower decline in sleep with aging ( ). It is unclear what accounts for these paradoxical findings between self-report and PSG sleep studies. It is possible that sleep time (percentage or total time), sleep onset, number of arousals, and SWS/SWA are not sensitive enough measures of poor sleep. More in depth studies of sleep intensity as measured via quantitative EEG or spectral analysis are needed. The sleep complaints commonly reported in women make up the core symptomology of the sleep disorder insomnia. By definition, insomnia is persistent difficulties initiating and maintaining sleep associated with poor sleep quality. Along 21

37 with changes in sleep onset, sleep time, and the number of night arousals, insomniacs have an increase in the higher frequency power, specifically in sigma and beta power during NREMS ( ). The presence of these higher frequency bands associated with cortical activation may be a better marker of poor sleep quality or insufficient sleep. Not surprisingly, there is a gender bias in insomnia, with women at a two-fold greater risk for developing insomnia throughout their lifetime compared to men (177, ). This difference in risk emerges at puberty ( ) and increases with age. More recently, EEG spectral analyses found that women have increased high frequency power compared to men during NREMS (198). In fact, women have higher spectral density across all frequency bands compared to men (183,199,200). These higher frequency bands may be the marker for poorer sleep in women that previous studies did not account for in their PSG analysis. Sleep debt appears to accumulate more quickly in women and may have more debilitating effects on overall health. Women experience greater SWA rebound compared to men, following a period of sleep deprivation (201). These data suggest that sleep may be regulated differently in men and women and that the effects of sleep loss may be more severe in one sex over the other. Insomnia and sleep disturbances are risk factors for the development of depression (191,202,203). Therefore, women are at even greater risk for affective disorders like depression and anxiety due to the higher prevalence of insomnia and sleep disruption throughout the lifespan (202,204,205). Emerging clinical findings indicate that women suffering from insufficient sleep are at a greater risk of metabolic and cardiovascular dysfunction ( ), affective mood disorders [for review (211)], and immune function dysregulation compared to men (212). 22

38 Hormonal Influences on Sleep Across the Lifespan Gonadal steroids are potent modulators of behavior. Gonadal steroids are steroid hormones synthesized from cholesterol by the gonads (testes in males, ovaries in females), and include testosterone, estradiol, and progesterone. In women, sleep complaints typically coincide with periods of hormonal fluctuation like during puberty, across the menstrual cycle, during pregnancy or across the menopausal transition [for review (213)]. Following menopause, sleep disruptions persist but they are alleviated by hormone replacement. This together with the fact that the onset of the sex difference and risk for insomnia at puberty implicates a role for the gonadal steroids, estradiol and progesterone, in the regulation of sleep in women. Women experience the greatest or most marked changes during times of hormonal fluctuation with menstrual cycle and menopause being the best-characterized periods of sleep disruptions in women. Less is known about the hormonal changes during puberty and pregnancy and their association with sleep disruptions. Hormonal changes at the onset of puberty are associated with an increased prevalence of sleep disruptions and prompt the emergence of sex differences in insomnia (197). During pregnancy, women experience significant changes in sleep [for review (214)]; however, it is difficult to parse out the direct effects of hormonal changes or those caused by physiological changes due to the growth and development of the fetus. As early as the first trimester, women experience increased fatigue and report poor sleep quality and restless sleep ( ). Studies indicate that initially total sleep is increased but then declines throughout the course of the pregnancy (216,217,219). Pregnancy and puberty, though significant, are relatively short periods of time in a woman s life when gonadal steroids can affect sleep. 23

39 Nonetheless, sleep loss during these periods may have long-term adverse effects on sleep. It is unclear if hormonal events during puberty or pregnancy can influence the sleep circuitry or its response to gonadal steroids or challenge long term. It is possible that sleep insufficiency during these periods can affect endocrine functioning long term, which can then negatively affect sleep and woman s overall health. Sleep Across the Menstrual Cycle From the onset of puberty until menopause, women experience fluctuations in gonadal steroids monthly across the menstrual cycle. In the earlier follicular phase, estradiol levels begin to rise, ultimately reaching a peak at mid-cycle which triggers ovulation (Fig. 4). Following ovulation, estradiol levels begin to decline but remain elevated compared to the early follicular phase. The post-ovulatory phase of the menstrual cycle is referred to as the luteal phase during which time progesterone levels reach a peak (Fig. 4). Studies investigating sleep disruptions throughout the menstrual cycle report mixed findings as to the extent of sleep disruptions. Typically, sleep complaints in women are commonly reported during the luteal phase, when estradiol and progesterone are high compared to during the follicular phase, when levels are low ( ). PSG studies only report increased NREMS spindle activity (224,225) and reduced REMS during the luteal phase (224, ). These studies highlight the inconsistency between subjective and objective measuring of sleep in women. Exogenous hormones, like oral contraceptive use, also influence sleep in women. PSG studies find that women taking oral contraceptives have more stage N2 NREMS 24

40 Figure 4. Fluctuation in ovarian hormonal milieu across the menstrual cycle. During the follicular phase, progesterone is at low level, whereas estradiol begins to rise mid-phase and peaks just before ovulation. Estradiol dips during ovulation but then remains at a steady elevated level during the luteal phase. Progesterone begins to rise just after ovulation and reaches a peak midluteal phase before dropping at the end of the phase. 25

41 compared to their placebo phase and naturally cycling women (230). SWS sleep is reduced (228,230,231) and REMS is increased (231). It remains unclear, however, which ovarian hormones are directly affecting sleep. Women suffering from menstrual-related disorders, like premenstrual dysphoric disorder (PMDD) and polycystic ovarian syndrome, experience significant sleep disruptions as well. Women with PMDD, for example, commonly suffer from insomnia and complain of poor sleep quality and frequent night awakenings (229,232,233). PSG studies found few changes in objective sleep parameters. One study found that women with PMDD have more stage 2 NREMS compared to controls but less REMS (229). Other studies found increased SWS in women with PMDD (234,235), while another did not find differences in objective sleep parameters (233). Subjective sleep parameters in women are not congruent with objective measures, providing further evidence that the parameters studied may not be indicative of sleep problems in women. EEG spectral analyses especially focusing on the higher frequency range may reveal more objective changes in sleep during these periods of poor subjective sleep in women. Sleep Across the Menopausal Transition As a woman ages, she undergoes changes in hypothalamic and ovarian function that leads to the cessation of reproductive capacity. The Stages of Reproductive Aging Workshop defined the stages of change (236). During the late reproductive stage, there are decreases in fecundity and changes in menstrual cycles may become noticeable. When a woman enters the menopausal transition, which is often referred to as perimenopause, menstrual cycle lengths become variable and hormones along the 26

42 reproductive axis begin to change. Amenorrhea longer than 60 days begins to occur and cycles occur with increasing variability. There are also extreme fluctuations in hormone levels and ovulation becomes sparse. Menopause is considered complete when menstruation ends and does not reappear for a year. Estradiol levels decline since the ovarian functioning ceased. During the postmenopausal period, estradiol levels and other hormones remain at low constant levels. Along with the physiological changes in hormones, perimenopausal and postmenopausal women experience vasomotor symptoms, like hot flushes and night sweats, sleep disturbances, changes in mood and affect, and changes in sex drive and function (237). 44% to 61% of peri- and postmenopausal women report symptoms of insomnia, including difficulties initiating and maintaining sleep [for review (213)]. Postmenopausal women have more frequent night awakenings, restless and poor quality sleep compared to premenopausal women ( ). Subjective sleep complaints do not correspond to objective markers of sleep during menopause. One study found that postmenopausal women sleep longer and have more SWS compared to premenopausal women (243). The disparity between subjective sleep complaints and objectively measured sleep in peri- and postmenopausal women is similar to that observed in younger and reproductive age women. A more recent study found that beta power in the EEG is elevated in peri- and postmenopausal women compared to premenopausal women (236). These findings suggest more cortical arousal during sleep during and after menopause. Hot flushes affect 75-85% of women across the menopausal transition. Hot flushes are often associated with subjective sleep disruption in menopausal women 27

43 (239,242, ); however, the exact relationship between them remains controversial. Some studies indicate that hot flushes are not associated with objective measures of sleep disruption ( ). However, pharmacologically induced hot flushes in young, nonmenopausal women correlates directly with the degree of sleep fragmentation and increased light sleep (253). These data suggest that there is a link between hot flushes, night awakenings, and poor sleep, but the nature of this link needs to be further explored. The effects of hormone replacement therapy during menopause on sleep disruptions are inconsistent. Many studies have found that hormone replacement therapy improves sleep quality and menopausal symptoms, including hot flushes ( ). Other studies indicate that estradiol replacement (with and without progesterone) alleviates vasomotor systems (hot flush) but does not significantly improve any sleep parameter in post-menopausal women ( ). Antonijevic and colleagues found that estradiol replacement improves subjective sleep quality and EEG but not objective measures of sleep (262). Inconsistencies in these findings may be due to differences in hormone replacement (type and duration) or the groups of women studied (age, reason for and duration of hypogonadism). Hormone-related Sleep Changes in Men It is unclear if gonadal steroids, mainly testosterone, affect sleep in men. Testosterone levels closely correspond to the sleep cycle. Testosterone is secreted during sleep and peak levels occur just before or after REMS onset (263,264). Sleep disruption decrease the amount of circulating testosterone (265,266) and lower levels of testosterone, which decline naturally with age, are correlated with less consolidated sleep 28

44 (267). Sleep apnea, a sleep disorder where there are pauses or disruptions in breathing during sleep, is common in men. Some studies found that testosterone influences the sleep apnea condition. High-dose testosterone replacement in older men decreases sleep efficiency and total sleep time and worsens sleep apnea (268). Others have also shown that testosterone administration is associated with sleep apnea ( ); however, blocking androgen action, via flutamide administration, does not affect sleep architecture or breathing parameters in men with sleep apnea (272). The exact relationship between sleep and testosterone remains unclear. Androgen deprivation therapy for prostate cancer is associated with high reports sleep disruptions [for review (273)]. Men undergoing this therapy frequently complain of hot flushes and night sweats that disturb their sleep ( ). Though uncommon, estradiol therapy can be used to treat the hot flushes and sleep disruption, as it has been shown to effectively alleviate hot flush in men and thereby improves sleep quality ( ). It is unclear through if estradiol has any effects on sleep in men. Sleep studies regarding the effects of gonadal steroids in both men and women have been rather inconsistent in their findings. The use of animal models is critical for furthering our understanding of how gonadal steroids affect sleep. We may clarify some discrepancies in our clinical findings if we better understand the circuits gonadal steroids can modulate and how they may impact sleep behavior. Rodent Models of Sleep Animal models are extensively used in sleep research. The rodent s neurocircuitry and neurochemistry of sleep share similarities with humans. Unlike humans, rats and 29

45 mice have less consolidated vigilance states (Fig. 5). They cycle through bouts of sleep (both NREMS and REMS) and wake during both the dark (active) and light (quiescent) phases. A higher percentage of sleep occurs in the light phase where as more consolidated bouts of wake occurring in the dark phase. In humans, sleep occurs during one consolidated period at night (Fig. 2A). Historian A. Roger Ekirch found that this was not always the case (280). Before the advent of artificial light, humans had segmented sleep consisting of two sleep periods. Technological advances enable us to work and live beyond sunrise and sunset forcing our sleep behavior to evolve. Rodents also proved to be excellent models for studying sex differences and hormonal modulation of sleep. Male rats and mice are typically used in sleep studies to avoid complications of the estrous cycle in females. But like humans, there are sex differences and hormonal modulation of sleep in both rats and mice. It is unclear what drives these differences. Sex Differences in Rodent Sleep Few comparative sleep studies in male and female rodents have been conducted ( ), therefore; our understanding of the underlying cause of sex differences in sleep is still unclear. In general, female rodents sleep less than male rodents ( ). Total sleep and NREMS are significantly lower in female mice compared to males (281,283,284) but REMS is significantly less in female rats compared to males (285,286). Sleep studies in mice indicate that females have more consolidated sleep-wake patterning than males with less state transitions, fewer arousals (or awakenings), and longer bout durations (281). 30

46 Figure 5. Representative hypnograms of rodent sleep-wake behavior. Wake (W), NREMS (N), and REMS (R) are represented throughout both phases, but rodents acquire more sleep during the (A) light phase and have consolidated periods of W during the (B) dark phase. 31

47 Sleep pressure is also different between male and female rodents. Data suggest that sleep pressure is greater in females compared to males. NREMS delta power, a quantitative measure of sleep pressure, is higher in females (281,287,288) and the percent increase in NREMS following sleep deprivation is greater in females. (281). Sex differences in sleep behavior and EEG power are eliminated following gonadectomy (GDX), suggesting that these differences are dependent on gonadal steroid hormones. Hormonal Modulation of Sleep in Rodents In rodents, gonadal steroid hormones exert robust effects on sleep-wake behavior. Fluctuations in the ovarian hormonal milieu across the estrous cycle correlate with changes in sleep and wakefulness in female rodents (285, ). On the night of proestrus, when the ovarian steroids estradiol and progesterone are elevated, both NREMS and REMS are significantly reduced compared to other phases of the estrous cycle (Fig. 6). Few studies found REMS rebound following proestrus (289,291,295). Ovariectomy (OVX) eliminates these fluctuations in time spent asleep or awake typically observed throughout the estrous cycle and hormone replacement (estradiol and progesterone or estradiol alone) is sufficient to recapitulate the suppression of sleep observed during proestrus in OVX rats (286, ) and to a lesser extend in mice (281,302). In aged OVX females, treatment with estradiol and progesterone or estradiol alone significantly increases wake and suppresses NREMS sleep compared to controls (303). Unlike in younger females, treatment of aged females with estradiol fragments sleep, such that there is a higher number of awakenings from NREMS compared to oil (303). 32

48 Figure 6. Fluctuation in ovarian hormonal milieu across the estrous cycle. Estradiol levels rise during the end of diestrus and peaks during proestrus. Progesterone levels rise on proestrus and peak toward the end of proestrus and beginning of estrus. 33

49 Sleep and wake behaviors in male rodents are insensitive to changes in gonadal steroids. Orchidectomy (ORDX) does not affect sleep or wake duration (281,304). However, NREMS is increased in ORDX mice treated with testosterone compared ORDX controls (302). A recent finding in adult male rats suggests that chronic estradiol replacement (via silastic capsule implants) induces arousal at the expense of sleep (304). The magnitude of change in males induced by chronic estradiol exposure compared to females is uncertain as only males were examined in that particular study. It is clear that gonadal steroids influence sleep, especially in females, but there are still significant gaps in our understanding related to how and why these changes occur. In 1990, Matsushima and colleagues implanted crystalline estradiol benzoate directly into the 3 rd ventricle and observed similar suppression of sleep to systemic estradiol (305). These findings indicate that estradiol is acting centrally to suppress sleep in female rodents. Influences of Sex and Hormones on the Sleep Sex difference in sleep may be caused by underlying genotypic differences between males and females or due to organizational and activational effects of gonadal steroids. Genotypic differences result from males having XY sex chromosomes and females having XX sex chromosomes. Sex chromosomes influence whether a male- or female-like phenotype develops. The presence of the SRY gene on the Y-chromosome leads to the expression of testis-determining factor and development of the testes (306). It is well established that genetic sex influences the development of the brain [for review ( )]. Beginning around embryonic day 18 until postnatal day 10, the 34

50 processes of masculinization, defeminization, and feminization occur [for review (307)]. These processes organize the neural substrate so that sex-specific behaviors occur when exposed to gonadal steroids in adulthood (Fig. 7). This defines the Organizational/Activational hypothesis of hormone action originally described by Phoenix, Goy, Gerall, and Young in 1959 (311). Their seminal work in guinea pigs showed that androgens (or their metabolites) during development permanently organize, or alter, the substrates mediating sexual behavior (311). Since 1959, our understanding of the mechanisms underlying sexual differentiation of the rodent brain has grown. In males, there is a surge in testosterone perinatally, around E18. The testosterone secreted during this period is aromatized into estradiol and it is the estradiol that organizes the neural substrates by influencing neuronal survival, dendritic branching and spines, axonal projections, and glial environment (312) [for review (308)]. Since the critical period extends into early postnatal life, sexual differentiation of the rodent brain can be manipulated. Males can be feminized by neonatal castration blocking estradiol s action or by preventing the conversion of testosterone to estradiol ( ). Conversely, females can be masculinized by exogenous administration of testosterone or estradiol (311, ). Postnatal manipulation of sexual differentiation of the brain is a valuable tool for studying sex differences in brain and in behavior. It is unclear if genetic or gonadal sex affects sleep-wake circuitry, as the majority of studies investigating the neurobiological basis of sleep has been done in males. It is possible that the sleep circuitry is organized differently during masculinization or feminization of the brain, which may lead to differences in sleep amongst the sexes. 35

51 Figure 7. The organizational/activational hypothesis of hormone action. Phoenix et al. determined that androgens (or their metabolites) during development permanently organize, or alter, the substrates mediating sexual behavior (311). Organization occurs during the critical period of development (~E18-P10). During this time, the processes of masculinization and defeminization occur to get a male brain. The presence of the Y-chromosome leads to the development of the testes, which releases testosterone during the critical period. Testosterone is aromatized to estradiol and it is the estradiol that organizes the neural substrate. In females, feminization occurs in the absence of testosterone/estradiol during the critical period. In adulthood, circulating levels of gonadal steroids activate the neural circuits organized during development. If masculinization and defeminization occurred then males will exhibit male-typical behaviors. Feminization of the brain will leads to female-like behaviors during adulthood. 36

52 Steroid Action Gonadal steroids have multiple mechanisms of action in target tissue. These actions are classified as either genomic or non-genomic effects [for review (319,320)]. Genomic effects are initiated when the steroid hormone binds its receptor and homodimerizes with another activated receptor. The homodimers translocate to the nucleus where it binds to steroid response elements within promoter gene sequences or interacts with other transcription factors to influence transcription of responsive genes. These effects are typically slow acting, taking hours to days to occur. Non-genomic effects can be mediated through interactions with kinase cascades, intracellular signaling pathway, and interactions with neurotransmitter receptors to have fast acting effects on the target cell. Both genomic and non-genomic effects of estradiol are generally initiated by estradiol binding one of its receptors (Fig. 8). There are two classical estrogen receptors (ER), ER alpha and ER beta. These receptors were originally thought to just act as transcription factors, but ER alpha and beta can also be tethered to the membrane and in a position to interact with intracellular signaling cascades ultimately leading to the phosphorylation of proteins and possibly changes in transcription. GPR30 is another receptor that binds estradiol. It is an orphan G-protein coupled receptor, localized to the membrane of the endoplamic reticulum (321). GPR30 activation leads to increased intracellular calcium and transduction of signaling cascades. Estradiol also has receptorindependent actions by acting as allosteric modulators of neurotransmitter receptors [for review (319)]. 37

53 Figure 8. Estradiol has genomic and non-genomic mechanisms of action. Genomic actions occur when estradiol (red circle) binds to its receptor in (ER; purple oval) in the cytoplasm of the cell. Once bound, activated ERs will homodimerize and translocate to the nucleus where they will either bind to estrogen response elements (ERE) or interact with other transcription factors (TF) to influence transcription of estrogen responsive genes. Estradiol can also have non-genomic actions by activating ER along the membrane that interacts with signaling cascades leading to phosphorylation of proteins. GPR30 is a G-protein coupled receptor on the membrane of the endoplasmic reticulum. Estradiol binding causes increased calcium in the cytoplasm, which can then activate signaling cascades. Lastly, estradiol can also bind to neurotransmitter receptors to modify their cellular signaling. Adapted from Stice and Knowlton [2008] Molecular Medicine. 38

54 Targets for Estradiol Action Estradiol may have direct actions within the sleep-wake circuitry. Many rodent studies indicate that wake-nuclei are target sites for gonadal steroids. Brainstem nuclei, specifically the LC and DRN contain ER alpha and beta and as well as androgen receptor (AR) expressing cells ( ). It is less clear if sleep-active nuclei contain ER or AR. The POA contains ER mrna expressing cells (323) and the MnPN contains ER protein, at least during early development in females (328). It is possible then that estradiol can act directly within key sleep-active areas to suppress sleep. Indirect evidence suggests a role for estradiol in the VLPO. Our lab has previously shown that there are both sex differences and hormonal modulation of VLPO activation. Fos expression in the VLPO is greater in females compared to males in the absence of circulating hormones (293). Fos expression in the VLPO is lower in females treated with estradiol compared to those treated with oil and Fos expression is unaffected by changes in circulating testosterone levels (293). Conversely, Fos expression is increased in the TMN in females following estradiol treatment (293). Deurveilher and colleagues also found increased Fos expression in the TMN, the laterodorsal subnucleus of the bed nucleus of stria terminalis and PF/LH (329). Our lab also discovered that lipocalin-type prostaglandin D synthase (L-PGDS), the enzyme responsible for the synthesis of the potent somnogen (sleep-promoting factor) prostaglandin D2, is also reduced following estradiol treatment (293). Similarly, in mice, others have shown that estradiol reduces mrna expression of L-PGDS in the VLPO and mrna expression of adenosine 2A receptors, a receptor critical for sleep induction by the somnogen adenosine, in the POA and VLPO are reduced following estradiol treatment in OVX mice 39

55 (330). It is plausible then that estradiol is actively reducing sleep-promoting signals within the POA to suppress sleep. The MnPN and VLPO are potential sites of estradiol action on REMS homeostasis. Estradiol modulates REMS homeostasis in a time-dependent manner. Following a period of sleep deprivation, estradiol treatment attenuates REMS rebound if females are allowed to recover in the dark phase (300,301), but enhances REMS rebound if they recover in the light (301). Neurons within the MnPN and VLPO were identified as sites for REMS homeostasis (91); therefore, it is possible then that estradiol has direct actions within the VLPO and MnPN to influence the homeostatic regulation of REMS. Based on these findings, we have targets within the sleep circuitry for exploring estradiol s actions on sleep. Using our established hormone replacement and sleep analysis paradigm, we can address the unanswered questions of how and why there are sex differences in sleep and how estradiol suppresses sleep in females in our model. Because most sleep studies are done in men or male rodents, treatment generalized to the male physiology may not effectively alleviate sleep disruptions in women. Advancing our understanding of the mechanisms underlying hormonal modulation of sleep is imperative for better treatment of sleep disruption in women. 40

56 Chapter 2: Goals of the Dissertation Since the early 1950s, studies investigating the influence of gonadal steroid hormones on sleep have focused on behavioral endpoints. Because of this, many gaps in our understanding still remain. While it is quite clear that gonadal steroid hormones, especially estradiol, suppress sleep in female rats and mice, what remains unclear is how estradiol suppresses sleep and why this sleep suppression seems to be sex-specific? Based on the literature and preliminary data in the lab, we propose the following hypothesis: Hypothesis: Estradiol suppresses activation of sleep-promoting neurons in the preoptic area to suppress sleep in a sex-specific manner. Prediction 1: Sex differences in sleep-wake behavior are due to activational effects of gonadal steroids on organized brain circuitry. Prediction 2: Estrogenic action in the preoptic area is required for estradiol s suppression of sleep in females. Prediction 3: Estradiol suppresses sleep by increasing downstream arousal signals like the orexinergic system. 41

57 Chapter 3: General Methods Animals All experiments were performed in Sprague-Dawley female ( g) and male ( g) rats. For developmental studies, adult females were mated in our animal facility to allow for manipulations of the pups immediately following birth when necessary. Animals for non-developmental studies were ordered from Charles River Laboratories. Due to a limited number of transmitters, recording slots, and/or animals per litter, all experiments were run in multiple cohorts, containing a minimum of two experimental groups per cohort. Animals were housed under a 12:12 light:dark or reversed cycle with free access to food and water for the duration of the study. The timing of treatments were determined based on zeitgeber time (ZT; lights on = ZT0), not the 24h clock time, to keep paradigms as similar as possible when light cycles switched between normal and revered. All procedures were performed in accordance with the NIH guide for care and use of laboratory animals. All experiments were approved by and were in accordance with the guidelines of the University of Maryland Institutional Animal Care and Use Committee. Surgeries All surgical procedures followed aseptic technique and were performed under isoflurane anesthesia. Prior to the start of surgery, pressure was applied to the animal s hind paws to confirm it was non-responsive to stimuli. 42

58 In all studies, adult female rats were ovariectomized (OVX), while some males were orchidectomized (ORDX) and some remained gonadally intact. For each experiment, the gonadal status of each animal will be noted. The flanks of each female rat were shaved and incisions were made into the skin and muscle wall. The ovary was located, ligated and removed. Sutures were used to close the muscle wall and the skin was closed using wound clips. This process was repeated on the opposite side of the animal. For male ORDX, a 1.5cm incision was made into the scrotum followed by a small incision into the muscle. The blood supply to each testis was ligated and both testes were removed. The muscle wall and skin were both sutured closed. Immediately following gonadectomy (GDX), each animal was implanted with TL11M2-F40-EET transmitters (Data Sciences International, MI, USA). A bi-potentiallead transmitter was implanted subcutaneously through a 3.0cm dorsal incision of the abdominal region. Another incision, approximately 4.0cm, was made along the midline of the head and neck to expose the skull and neck muscle. Two burr holes were drilled asymmetrically and stainless steel screws (Plastics One, Roanoke, VA) were implanted at 3.0mm anterior/ +1.5mm lateral and 7.0mm posterior/ -1.5mm lateral relative to bregma. The four transmitter leads were threaded subcutaneously to the opening of the head incision. Electroencephalographic (EEG) leads were wrapped around the screws and secured with dental cement. Electromyogram (EMG) leads were implanted directly in the dorsal cervical neck muscle, approximately 1.0mm apart, and sutured in place. The skin along the head was sutured closed while the dorsal incision was stapled closed with wound clips. All animals were treated with antibiotic ointment and topical lidocaine as 43

59 well as carpofen (5mg/kg) post-operatively, and then allowed 7 days to recover before the start of the experiments. Hormone Replacement Paradigm All experiments follow a within-animal design. All animals received an oil injection for baseline data recordings followed by two injections of gonadal steroids 24h apart (Fig. 9). Estradiol benzoate (EB; 5µg then 10µg; Sigma-Aldrich, St. Louis, MO) was used to exogenously replace estradiol. EB is a synthetic ester of estradiol. Nonspecific steroidal esterases remove the conjugated ester to produce biologically active estradiol in vivo (331,332). This replacement paradigm delivers physiological doses of estradiol that mimic the gradual rise in estradiol that occurs during proestrus (301). Testosterone propionate (TP; 500µg; Sigma-Aldrich, St. Louis, MO) was used to exogenously replace testosterone. The dosage of TP has been reported to reinstate physiological levels of testosterone in males within 24h (333). Figure 9. Gonadal steroid replacement paradigm. Arrows indicate injection times. Data Acquisition EEG and EMG data were collected using Dataquest ART 4.0 software (DSI). For each animal, data collected during the periods of interest were scored in 10-second epochs 44

60 with Neuroscore 2.0 (DSI). Each epoch was scored as wake (low-amplitude, high frequency EEG with high amplitude EMG), NREMS (high-amplitude, low frequency EEG with low amplitude EMG), or REMS (low amplitude, high frequency EEG with muscle atonia or periodic muscle twitches). Figure 1 shows representative EEG and EMG traces from Neuroscore for recordings scored as wake, NREMS, and REMS. Transitions into a new vigilance state were only considered if 2 epochs (20 sec) of the new vigilance state was observed. The total duration (in minutes) for each vigilance state was analyzed for the light and dark phase for each scoring period. Immunocytochemistry Animals were overdosed with a ketamine/acepromazine mix before being transcardially perfused with 0.9% sodium chloride + 2% sodium nitrite solution followed by 4% paraformaldehyde in 0.05M potassium phosphate buffered saline (KPBS). The brains were removed and postfixed overnight in 4% paraformaldehyde. Brains were cryoprotected in 30% sucrose in KPBS, frozen on dry ice, and stored at -80 o C. Each brain was cut on a cryostat along the coronal plane at 30µm thick into 4 series and stored in an ethylene glycol-based storage solution at -20 o C until processed. Sections in each series are separated by 120µm. Cohorts containing animals from all treatment groups were immunocytochemically processed in the same tray. Briefly, sections were rinsed free of the cryoprotectant in 0.05M KPBS, reacted with 1% sodium borohydride in KPBS and rinsed. Sections were then incubated in 1.0% hydrogen peroxide to release endogenous peroxidase activity, rinsed and incubated with specific primary antibodies for the protein of interest. After 45

61 primary incubation, sections were rinsed in KPBS and incubated in biotinylated secondary antibodies for 1h, followed by washes in KPBS. Finally, the samples were incubated with an avidin-biotin horseradish-peroxidase complex (Vectastain ABC, Elite Kit; Vector Laboratories) for 1h at room temperature then washed in 1X Tris-buffered saline (TBS). The sections were visualized with nickel sulfate (25 mg ml) and 3,3 - diaminobenzidine tetrahydrochloride (DAB; 0.2 mg ml; Polysciences, Warrington, PA, USA) in TBS solution containing 0.005% H 2 O 2, rinsed in TBS and transferred to KPBS. After visualization, the sections were mounted serially on 2% gelatin-coated glass slides and coverslipped. 46

62 Chapter 4: Neonatal Exposure to Gonadal Steroids Organizes Adult Sleep Behavior. Introduction Basic research studies show that female rodents sleep less than males ( ). In mice, total sleep and NREMS are significantly lower compared to males (281,283,284), whereas female rats have significantly less REMS than males (285,286). These differences are eliminated following GDX, suggesting that these sex differences are dependent on gonadal steroid hormones (281,285,286). It is unclear whether sex differences in sleep are due to gonadal steroid hormonespecific effects, i.e. estradiol and/or progesterone-mediated effects, or if underlying differences due to sexual differentiation of the brain contribute to these effects as well. Two early studies began investigating these questions. First, Branchey and colleagues feminized male rats by neonatal castration and found that it reduces NREMS and REMS in adulthood following a systemic injection of estradiol (334). Conversely, estradiol does not reduce REMS in intact adult male rats. Second, Yamaoka showed that the circadian NREMS and REMS rhythms in female rats masculinized at birth do not change following an injection of estradiol (286). Neither study, however, directly compare the effects of estradiol in feminized males or masculinized females to control females. Furthermore, it remains unclear what masculinization and/or feminization does to the sleep wake circuitry to lead to this sex-specific effect of estradiol on sleep. The VLPO is sensitive to changes in estradiol levels. Previously, we have shown that there is a sex difference in the hormonal modulation of both the activity and expression of enzymes responsible for producing known somnogens in the VLPO (293). There sex difference in gonadal steroid hormone mediated suppression of (a) sleep-active 47

63 neurons and (b) protein levels of lipocalin-type prostaglandin D synthase, the synthesizing enzyme for the somnogen prostaglandin D2. Female rats but not male rats respond to changes in the hormonal milieu (via estradiol replacement following OVX or castration, respectively). These differences may not only be dependent on gonadal steroid hormones but also on sexually differentiated circuitry. In order to address the origin of the sex differences observed in sleep, we systematically tested the basic principles defining sex differences in the brain as recently outlined by McCarthy et al. (310). These sex differences may be due to activational effects (transient actions following exposure to gonadal steroids), organizational effects (permanent changes in the neural substrate induced by exposure to gonadal steroids during perinatal development or puberty that can alter how the brain responds), or both. We hypothesize that sex differences in sleep are due to activational effects of estradiol on organized sleep-circuitry. To test this, we first GDX adult male and female rats and analyzed their sleep behavior following sex-specific steroidal manipulations. We also tested whether pubertal events in females sensitize their sleep behavior to estradiol s effects. Then, to test whether organization of the brain circuitry by early exposure to gonadal steroids contributes to sex differences in sleep, female pups were masculinized via neonatal exposure to testosterone. These manipulations allowed for a direct comparison in the magnitude of change following steroid replacement in males and females. Finally, we used Fos-immunoreactivity within the VLPO to determine whether sexual differentiation during early development organized a component of the sleep circuitry, a significant gap in our knowledge. 48

64 Methods Sleep Recordings and Analysis In adulthood, all animals were GDX and simultaneously implanted with TL11M2-F40-EET transmitters (Data Sciences International, MI, USA) under isoflurane. Briefly, immediately following GDX, a bi-potential-lead transmitter was implanted subcutaneously to allow for collection of EEG and EMG data. All animals were treated with antibiotic ointment and topical lidocaine as well as carpofen (5mg/kg) postoperatively, and then allowed 7 days to recover before the start of the experiments. The total duration (in minutes) for each vigilance state was analyzed for the light and dark phase for each scoring period. To compare the magnitude of change induced by gonadal steroid hormone replacement, the percent change from oil baseline for each gonadal steroid was calculated [((hormone-oil)/oil) *100]. Effects of Sex-Specific Gonadal Steroid Replacement on Sleep Female (n=8) and male (n=4) rats were used to determine if there are sex differences in the effects of sex-specific gonadal steroid replacement on sleep. A subset (n=4) of control females was OVX on postnatal (PN) day 22, prior to puberty. Vaginal opening, a commonly used marker of puberty, did not occur in any of these females, based on visual inspection. Prepubertal OVX effectively eliminates the onset of adult ovarian hormones and estrous cyclicity and thus was used to determine whether females responsivity to the modulatory effects of estradiol on sleep and wakefulness in adulthood is set during development or if further modification at the time of puberty is required. 49

65 All animals received an oil injection for baseline recording. Then, females received two injections, 24h apart of EB (5µg then 10µg), while males received two injections 500µg of TP, according to our standard paradigm (General Methods, Fig. 9). Sleep analysis was done on the 24h period 24h following the last gonadal steroid hormone injection. Effects of Brain Organization and Gonadal Steroids on Sleep A subset of female pups (n=5) received subcutaneous injections of 100µg of TP on the day of birth and PN day 1 to masculinize the neural substrates, as previously described (335,336). These females are referred to as masculinized females or throughout the manuscript and represent genetic females with male-like brain organization. Control female (n=6) and male littermates (n=7) were treated with 100µl sesame oil on the same treatment days. All animals were raised to adulthood (~80 days old) and used for the sleep studies described below. Neonatal androgen treatment results in an anovulatory condition characterized by lack of estrous cyclicity and reduced ovarian weight that has been attributed to a reduced numbers of ovulatory follicles ( ,337), traits indicative of a masculinized hypothalamic-pituitary-gonadal axis. Paired ovarian weights were recorded at the time of adult ovariectomies, as an independent measure of the masculinizing effects of neonatal TP exposure in females. These control females and masculinized females were used in the Fos-immunoreactivity experiment described below. Neonatal TP exposure significantly reduced the paired ovarian weight by 31% compared to females treated with 50

66 oil (two tailed t-test; t 10 = 3.96, P < 0.003; mean weights ± SEM in grams were: masculinized females = ± 0.01 and females = ± 0.01). These cohorts of animals were used to test whether exposure to gonadal steroids neonatally organizes the sleep circuitry in such a way, which alters the effects of estradiol and testosterone on sleep behavior. The order of hormone replacement (either EB or TP) was randomized in each group to control for timing. The second round of injections began 7 days following the end of the first recording period to ensure that gonadal steroids were no longer in circulation. Androgen s Effect on Sleep Dihydrotestosterone (DHT) benzoate (DHTB; Sigma-Aldrich, St. Louis, MO, USA), a non-aromatizable androgen was administered to GDX male and female rats (n=4 per group) to determine whether testosterone exerts its effects through androgens or via aromatization to estradiol. Similarly to the EB and TP injection paradigm, two injections of 500 µg DHTB were given 24h apart (General Methods, Fig. 9). This dose of DHTB has been reported to effectively restore copulation in GDX males (333). Fos-immunoreactivity (Fos-ir) In a separate cohort of masculinized females (n=5), control females (n=6), and males (n=6), animals were overdosed using ketamine/acepromazine mix before being transcardially perfused with 0.1 M phosphate-buffered saline (PBS) followed by 4% paraformaldehyde in PBS. The brains were removed, post-fixed in 2.5% acrolein / 4% paraformaldehyde in PBS; cryoprotected in 30% sucrose in PBS; frozen on dry ice and 51

67 stored at -80ºC before being sectioned in a cryostat into 30µm thick sections along the coronal plane. The sections were placed into an ethylene glycol-based storage solution at -20ºC until processed for Fos-ir. Cohorts containing animals from all treatment groups were immunocytochemically processed in the same tray. Sections were processed for Fos-ir following our standard protocol (General Methods). Sections were incubated for 48h at 4 C with rabbit polyclonal Fos antibodies (Ab 4191; Oncogene Science, MA, USA) at a dilution of 1: in 10% normal goat serum and 0.3% Triton X-100 in PBS. Sections were visualized with a nickel sulfate-enhanced DAB in sodium acetate solution containing 0.005% H 2 O 2, rinsed in acetate solution and transferred to PBS. After visualization, the sections were mounted serially on 2% gelatin-coated glass slides and coverslipped. Fos-ir Quantification and Analysis We systematically counted the number of Fos-ir cells in the VLPO with the aid of the Neurolucida software, an image-combining computer microscopy program (MicroBrightField, Colchester, VT, USA) as previously described (293). Briefly, slides were anatomically matched and numerically coded so that the investigator conducting the analysis was blind to the experimental group. Sections analyzed were from one-in-three series (adjacent sections were separated by 90µm). Three brain sections corresponding to Plates 19 and 20 of the Paxinos and Watson rat brain atlas (90) were used in the analysis. The placement and size of the contours were in accordance with previously defined parameters (338). The contour began ~1mm from the ipsilateral ventricular wall and 52

68 extended 0.7mm laterally and 0.3mm dorsally. Six contours per bilateral section of the VLPO were counted and an average count of Fos-ir was derived. Estrogen receptor alpha -ir (ER alpha-ir) In a separate cohort of animals, females and males were collected 5h into the light phase and 5h into the dark phase. Females were treated with either oil or estradiol and all males were gonadally intact. Groups were collapsed and analyzed for sex differences in ER alpha-ir expression in the MnPN and VLPO (female n=16; males n=6). All animals were overdosed with overdosed with a ketamine/acepromazine mix before being transcardially perfused with 0.1M KPBS followed by 4% paraformaldehyde in KPBS. The brains were removed, post-fixed, cryoprotected, frozen on dry ice and stored at -80ºC before being sectioned in a cryostat into 30µm-thick sections along the coronal plane. The sections were placed into an ethylene glycol-based storage solution at -20ºC until processed for ER alpha-ir. Sections were processed for ER alpha-ir following our standard protocol (General Methods). Sections were incubated for 48h at 4 C with rabbit polyclonal anti-estrogen receptor alpha antibodies (06-935; EMD Millipore, MA, USA) at a dilution of 1:50,000 in 5% normal goat serum and 0.05% Triton X-100 in KPBS. Sections were visualized with a nickel sulfate-enhanced DAB (Polysciences, Warrington, PA, USA) in sodium acetate solution containing 0.005% H 2 O 2, rinsed in acetate solution and transferred to PBS. After visualization, the sections were mounted serially on 2% gelatin-coated glass slides and coverslipped. 53

69 ER alpha-ir Quantification and Analysis Slides were anatomically matched and numerically coded so that the investigator conducting the analysis was blind to the experimental group. Sections analyzed were from two-in-four series (adjacent sections were separated by 60µm). Six brain sections corresponding to Plates 33 to 35 of the Paxinos and Watson rat brain atlas (90). The Optical Fractionator workflow in StereoInvestigator (MicroBrightField, Colchester, VT, USA) was used to determine an estimated population of ER alpha-ir cells in MnPN. A grid was placed within MnPN contours measuring 100 x 100µm and ER alpha-ir cells were counted within 50 x 50µm counting frames. The total estimated population calculated using mean thickness of the sections was used for each animal. ER alpha expression in the VLPO was too low for accurate estimates using stereological technique. Instead, all ER alpha positive cells were counted in three sections from one-in-four series (adjacent sections were separated by 120µm). Sections corresponding to Plates 19 and 20 of the Paxinos and Watson rat brain atlas (90) were used in the analysis. The placement and size of the VLPO contours were in accordance with previously defined parameters (338). The contour began ~1mm from the ipsilateral ventricular wall and extended 0.6mm laterally and 0.3mm dorsally. Three contours per bilateral section of the VLPO were counted and an average count of ER alpha-ir per section was derived. Statistical Analysis Student s t-tests were used to determine if there were differences in the total time females and males spent in wake, total sleep, NREMS, and REMS under baseline and 54

70 steroid hormone replaced conditions. Two-way repeated measures ANOVAs were run on the mean times spent in each vigilance state between GDX, oil treated males and females and steroid hormone replaced males and females during the light and dark phase of the 24h baseline day. Data were binned into 3h blocks across the light:dark cycle. Paired t-tests were used to determine whether estradiol in females and testosterone in males alter the mean time in wake, NREM or REM sleep in both phases compared to their oil-baselines. Mann-Whitney U nonparametric tests were used to compare the percent change of a vigilance state induced by sex-specific gonadal steroids between males and females. Mann-Whitney U nonparametric tests were also used to determine whether undergoing puberty is required for estradiol s effects on sleep in adulthood The mean times in each vigilance state were analyzed using two-way repeated measures ANOVAs followed by Bonferroni post hoc tests with the adult hormone status (oil baseline vs. gonadal steroid replacement) and brain organization (feminization vs. masculinization) as independent variables. A Kruskal-Wallis test was used to compare the percent change between females, males and masculinized females for each steroid hormone treatment. One-way ANOVA was performed on the mean number of Fos-ir cells in the VLPO and a Tukey s Multiple Comparison post hoc test was used to determine significance between groups. Student s t-tests were run to determine if there are sex differences in ER alpha-ir expression in the MnPN and VLPO. 55

71 Results Effects of sex-specific gonadal steroid replacement on sleep Quantitative analysis of the EEG/EMG traces demonstrated no significant differences in the total time spent in wake, total sleep, NREMS, or REMS in both the 12h light (wake: t 6 = 0.28, P > 0.05; total sleep: t 6 = 0.20, P > 0.05; NREMS: t 6 = 0.43, P > 0.05; REMS: t 6 = 0.38, P > 0.05) and dark (wake: t 6 = 0.31, P > 0.05; total sleep: t 6 = 0.20, P > 0.05; NREMS: t 6 = 0.37, P > 0.05; REMS: t 6 = 0.07, P > 0.05) phases between GDX females and males (Fig. 10; left panels). Since 12h totals can mask any significant changes that occur during specific times throughout the phase, we also looked at 3h bins across each phase. There were no significant differences in wake (F 1,42 = 0.11, P > 0.05), total sleep (F 1,42 = 0.09, P > 0.05) NREM sleep (F 1,42 = 0.17, P > 0.05) and REM sleep (F 1,42 = 0.31, P > 0.05 ) between GDX males and females in either the dark or light phase (Fig. 10, right panel). For each vigilance state, there was a main effect of time (wake: F 7,42 = P < , NREM: F 7,42 = P < , total sleep: F 1,42 = 60.75, P < , REM: F 7,42 = 8.34 P < ). Both males and females showed a circadian rhythm in sleep and wake behaviors. During the dark (or active) phase, both sexes are awake more, while they acquire more NREMS and REMS during the light (or quiescent) phase. From this point, sleep in the absence of gonadal steroids is referred to as baseline sleep. Estradiol and testosterone replacement in females and males, respectively, was used to examine the effects of gonadal steroids on time spent in each vigilance state. First, we found a difference in the amount of REMS between estradiol-treated females and testosterone-treated males during the dark phase (Fig. 11D, left). Following sex- 56

72 Figure 10. Sleep-wake behavior in GDX female and male rats. There are no differences in (A) wake, (B) total sleep, (C) NREMS, (D) or REMS during the light or dark phase in GDX female and male rats. Not 12h totals (left) nor 3h bins across the 24h day (right) revealed sex differences in any vigilance state. 57

73 specific gonadal steroid replacement, females acquire less REMS than males (t 6 = 2.40; P = 0.05). There is not a significant difference in 3h binned time (F 1,42 = 0.12, P > 0.05; Fig. 11D, right). Wake, total sleep, and NREMS did not differ between estradiol-treated females and testosterone-treated males in either the light (wake: t 6 = 0.64, P > 0.05; total sleep: t 6 = 0.64 P > 0.05; NREMS: t 6 = 0.24, P > 0.05) or dark (wake: t 6 = 1.50, P > 0.05; total sleep: t 6 = 1.50, P > 0.05; NREMS: t 6 = 1.20, P > 0.05) phase (Fig. 11A-C, left and right). REMS was not significantly different in the light phase (REMS: t 6 = 1.04, P > 0.05). Similar to our previous findings (293,300,301), estradiol in females increased wake at the expense of sleep, during the 12h dark phase (Fig. 12). On average, estradiol significantly increased wakefulness (t 3 = 11.51, P < 0.001) by about 12.5% or 63.5 minutes. Total sleep (t 3 = 11.25, P < 0.01), NREMS (t 3 = 6.698, P < 0.01) and REMS (t 3 =20.86, P < 0.001) are significantly suppressed by estradiol by about 30.7% (62.3 mins), 21% (33.7 mins) and 64% (28.6 mins), respectively. In males, testosterone did not significantly change any vigilance state (Fig. 12); however, there were trends for testosterone-mediated increased wakefulness (t 3 = 2.925, P = 0.06) and decreased total sleep (t 3 = 2.80, P = 0.07) REMS (t 3 = 3.039, P = 0.06). In the light phase, gonadal steroid replacement had fewer effects on behavior (Table 1). In females, there was a small but significant suppression of wake by estradiol (~17 mins; t 3 = 3.189, P < 0.05). There were no additional effects on either sleep phase or any significant effects of testosterone in males during the light phase. Analysis of the percent change from baseline for each vigilance state demonstrates a sex difference in the magnitude of change induced by the sex-specific 58

74 Figure 11. Sleep-wake behavior following sex-specific gonadal steroid replacement in female and male rats. There were no differences in 12h totals or 3h bins across time for (A) wake, (B) total sleep, or (C) NREMS, but estradiol-treated females had less REMS than testosterone-treated males in the 12h dark phase (D, left) but not across the 3h bins (right). 59

75 Figure 12. Effects of sex-specific gonadal steroid replacement on sleep-wake behavior in female and male rats. Estradiol significantly increased (A) wake and decreased (B) total sleep, (C) NREMS, and (D) REMS in female rats. In males, testosterone increased (A) wake and decreased (B) total sleep and (D) REMS but these effects did not reach statistical significance. *, P < 0.05 vs. oil. 60

76 Table 1. Effects of sex-specific gonadal steroid replacement on sleep-wake behaviors during the light phase Sex Wake (min) NREMS (min) REMS (min) Oil Sex-Steroid Oil Sex-Steroid Oil Sex-Steroid Female ± ± 11.3* ± ± ± ± 10.4 Male ± ± ± ± ± ± 8.1 Data are represented as mean ± SEM. *, P < 0.05 vs. oil. steroid replacement for total sleep (U = 0.00, P < 0.05; Fig. 13A) and REMS (U = 0.00, P < 0.05; Fig. 13A). Estradiol significantly suppressed total sleep in females by about 31% whereas testosterone suppressed sleep in males by about 12% compared to baseline. REMS in females was suppressed by about 64% while T only suppresses REMS in males by about 25%. Finally, neither the timing of the OVX (pre- or postpuberty) nor the timing of the first exposure to estradiol significantly affected the magnitude of estradiol s increase in wake or suppression of sleep (Fig. 13B). Estradiol and testosterone affect sleep and wake in females only Sleep and wake analysis following estradiol and testosterone replacement (randomized) in females, males, and masculinized females was used to determine whether brain organization (e.g., masculinized vs. feminized) underlies the responsivity of the sleep circuitry to the modulatory effects of gonadal steroids. All groups were compared back to their oil baseline following estradiol replacement (Fig. 14). There was a significant interaction between estradiol and brain organization for wake (F 2,15 = 5.277, P < 0.05) and REMS (F 2,15 = 5.708, P < 0.05). Estradiol in females significantly increased wakefulness (t = 3.406, P < 0.05) and decreased REMS (t = 5.507, P < 0.001); however, 61

77 Figure 13. Magnitude of change in vigilance states following gonadal steroid replacement. (A) Estradiol suppresses sleep in females significantly more than testosterone suppresses sleep in males. (B) Prepubertal OVX does not affect estradiol ability to suppress sleep in adulthood. *, P <

78 Figure 14. Effects of estradiol and testosterone replacement on sleep-wake behavior in females, males and masculinized females. (A) Estradiol increased wake at the expense of NREMS and REMS sleep in female rats. REMS was significantly reduced following testosteroneadministration in females, but the increased wake and suppressed NREMS did not reach statistical significance. (B) Males and (C) masculinized females were insensitive to the modulatory effects of estradiol and testosterone on sleep. *, P < 0.05 vs oil. 63

79 estradiol was ineffective at modulating sleep and wakefulness in animals with masculinized brain circuitry. There was a significant interaction between estradiol and brain organization for NREMS (F 2,15 = 3.967, P < 0.05); however, due to the stringency of the Bonferroni post hoc analysis and the increase in comparisons no significant effects were found. Since estradiol has been previously shown to suppress NREMS in females in our lab (293,300,301), an uncorrected Fisher s LSD test was applied and demonstrated that estradiol significantly suppressed NREMS (t = 2.196, P < 0.05). Next, all groups were compared back to their oil baseline following testosterone replacement (Fig. 14). Testosterone also had a significant main effect on REMS (F 1,15 = 10.60, P < 0.01), as it suppressed REMS in females (t = 3.819, P < 0.01) but not masculinized females or males. There were no significant effects of testosterone or brain organization on wake or NREMS. Analysis of the percent change in each vigilance state induced by gonadal steroid replacement further supports the notion that brain organization underlies the responsivity to the modulatory effects of gonadal steroids on sleep. The magnitude of estradiol s suppression of REMS is significantly different (H 2 = 7.630, P < 0.05; Fig. 15). Dunn s Multiple Comparison test revealed a significant difference between females and males (P < 0.05) but not between females and masculinized females or males and masculinized females. There was not a significant difference in the percent change in REMS induced by testosterone-replacement amongst the 3 groups (H 2 = 4.841, P > 0.05). The percent change induced by estradiol for wake and NREMS did not reach statistical significance but there was a trend for estradiol to increase wake in females compared to males and 64

80 Figure 15. The magnitude of change following estradiol and testosterone on sleep and wake in females, males and masculinized females. (A) There was a trend for a significant effect of brain organization on the magnitude of estradiol-induced arousal. (B) Estradiol and testosterone suppressed NREMS sleep in females but increased NREMS in males and masculinized females (did not reach statistical significance). (C) Estradiol suppresses REMS significantly more in females than in males. The magnitude of REMS suppresses was not different between masculinized females and females or males. There was a trend for testosterone suppressing REMS greater in females but that did not reach significance. 65

81 masculinized females (H 2 = 5.68, P = 0.06) and to decrease NREMS in females while it increases NREMS in males and masculinized females (H 2 = 5.42, P = 0.07). Fos expression in the VLPO is sensitive to the organizing effects of gonadal steroids. We showed previously that there is a sex difference in the number of sleep- active cells (via the expression of Fos) in the VLPO. Females have more Fos-ir cells at baseline than males (293). Here, Fos-ir cells in the VLPO of females, males, and masculinized females collected 8h into their sleep (light) phase were quantified to examine whether this sleep-associated brain nucleus is sensitive to the organizing effects of gonadal steroids. Females possessed 58% and 46% more Fos-ir cells compared to males and masculinized females, respectively (F 2,12 = 15.80, P < 0.01; Fig. 16). Androgen does not mediate changes in sleep and wakefulness. DHT, a non-aromatizable androgen, does not significantly affect wake, NREMS, or REMS in females or male, indicating that it is estradiol that suppresses sleep, not androgen. Data are presented in Table 2. ER alpha expression in the MnPN and VLPO. ER alpha is expressed in both the MnPN and VLPO but there is much less expression in the VLPO (Fig. 17A). In the MnPN, females have significantly more ER alpha-ir cells than males (t 20 = 3.42, P = 0.003; Fig. 17B). ER alpha expression in the VLPO was similar in females and males (t 20 = 0.82, P > 0.05; Fig. 17C). 66

82 Figure 16. Developmental organization of Fos expression in the VLPO. (Left) Photomicrographs of Fos-ir in the VLPO. (Right) Males and masculinized females have 58% and 46% less Fos-ir cells, respectively, compared to females. *, P < 0.05 vs. all other groups. 67

83 Figure 17. ER alpha expression in the MnPN and VLPO. (A) Representative photomicrographs of ER alpha-ir cells in the MnPN (left) and VLPO (right). The MnPN has significantly more ER alpha expression than the VLPO. (B) Representative photomicrographs of ER alpha-ir cells in the MnPN in females (left) and males (right). Females have significantly more ER alpha expression in the MnPN compared to males. (C) Representative photomicrographs of ER alpha-ir cells in the VLPO in females (left) and males (right). ER alpha expression in the VLPO is similar in females and males. *, P < 0.05 vs. females. 68

84 Table 2. Effects of DHT-replacement on the duration of sleep-wake behaviors females and males during the dark phase. Sex Wake (min) NREMS (min) REMS (min) Oil DHT Oil DHT Oil DHT Female ± ± ± ± ± ± 8.2 Male ± ± ± ± ± ± 5.6 Data are represented as mean ± SEM. *, P < 0.05 vs. oil. Discussion Here, the role of gonadal steroid hormones was tested in the organization and modulation of sleep in male and female rats. First, we found that exogenously administered estradiol has a greater magnitude of effect on sleep in females compared to males and masculinized females. Such a finding indicates a sex difference in the sensitively of the sleep circuitry to the modulatory effects of estradiol. Second, our data strongly indicate that estradiol suppresses sleep, and that androgens do not play a significant role. Finally, significant differences in Fos expression in the VLPO suggest that a component of the preoptic neural circuitry involved in sleep is organized by early exposure to gonadal steroids. Together, these data indicate that sex differences in sleep and wakefulness are due to activational effects of estradiol on organized brain circuitry. Previous studies raised the possibility that sexual differentiation of the brain played a significant role in the sex differences observed in sleep. Branchey et al. were able to suppress sleep in males who were feminized by neonatal ORDX but not in ORDX males in adulthood (334). Conversely, Yamaoka showed that masculinization of the female brain prevents estradiol from altering the circadian NREMS and REMS rhythms (286). Our findings support this early work and demonstrate that the critical period for brain sexual differentiation contributes to the foundations of sex differences in the 69

85 physiological processes mediating sleep. By directly comparing females, males and masculinized females under the same hormonal conditions, we determined that the sex difference in sleep is due to differences in the sensitivity of the circuitry to gonadal steroids. The VLPO is organized by early exposure to gonadal steroids. The sex difference in Fos expression in the VLPO was observed under baseline conditions when sleep behavior is the same between females and males. To our knowledge, this is the first report of clear evidence that a sleep-related nucleus is organized by early exposure to gonadal steroids. Organization of the VLPO can cause differences in cell number, type and connectivity. The lower levels of Fos expression in males and masculinized females may be due to the absence of neurons, not just the lack of activation, and it may be those neurons that are sensitive to estradiol, which may explain why estradiol only modulates sleep in females. Following the onset of puberty, which marks the activation of the hypothalamicpituitary-gonadal axis and gonadal maturation [for review (339)], women and female rodents experience changes in their sleep behavior along the menstrual and estrous cycle. Using a rat model, Sieck et al. assessed sleep and wakefulness before and after either natural or precociously induced puberty (340,341). Following their designated pubertal event, vaginal opening, there is an increase in wake at the expense of both NREM and REM sleep, suggesting that these changes in sleep and wake are mediated by the hormonal events occurring during puberty, not age-related changes (340,341). Until now, it was unclear whether the hormonal events during puberty were necessary for the adultlike suppress of sleep induced by estradiol. Our data suggest that the responsivity of the 70

86 circuitry underlying estradiol s effects on sleep in females is set early during development and that hormone-dependent changes that occur during puberty have no further effects on adult sleep organization. Unlike the robust and reproducible suppression of sleep in females induced by estradiol, our behavioral findings in males following gonadal steroid replacement were inconsistent. We initially found a trend towards increased wake and suppressed sleep in males in our first experiment investigating the effects of sex-specific gonadal steroids on sleep. This suppression of sleep by testosterone in males was not significant or reproducible (compare Fig. 12 and 14). The magnitude of the effect in males is relatively small and any statistically significant effect is tenuous at best because of variability in a cohort. A recent study showed that chronic estradiol replacement (via silastic capsule implants) significantly induces arousal at the expense of sleep in males (304). Nonetheless, as we demonstrated, differences between males and females are due to differences in the sensitivity of the circuitry and behavior to gonadal steroids. Here, we show that masculinization of the brain renders sleep behavior in males and masculinized females less sensitive to the suppressive effects of estradiol compared to females. Differences in ER expression may underlie differences in the sensitivity of the sleep circuitry and behavior. Estradiol has multiple mechanisms of action and many are mediated through binding its receptors (ER alpha or beta). The MnPN has abundant ER alpha expression of which there is a sex difference. Greater expression of ER alpha in the MnPN of females highlights this region as a potential target for estradiol s sex-specific effects on sleep. 71

87 Overall, the current findings implicate a significant role of sexual differentiation in the effects of gonadal steroids on sleep. Understanding why estradiol only affects sleep in females may help elucidate possible mechanisms of action. Sleep-circuits organized during develop are targets for isolating estradiol s effects on sleep and wake. 72

88 Chapter 5: The sleep-active preoptic area as a key site in estradiol s suppression of sleep Introduction As previously described in Chapter 4, sleep behavior in female rats is sensitive to changes in gonadal steroids, specifically estradiol, whereas sleep behavior in males is insensitive. Sexual differentiation of the brain during development plays a key role in the differences found in the hormonal modulation of sleep in rats. What remains unclear is where in the brain estradiol is acting to suppress sleep in females. Sleep-wake behaviors are regulated by reciprocal connections between sleeppromoting nuclei in the POA (MnPN and VLPO), and arousal centers in the hypothalamus and brainstem [for review (95)]. The VLPO and MnPN are two key sleepactive nuclei involved in the onset and maintenance of sleep (83,86,88,90). Previously, we found ER alpha protein expression in both the MnPN and VLPO, but expression was markedly greater in the MnPN. We also found a sex difference in ER alpha expression in the MnPN. Females had higher levels of expression compared to males. These expression patterns of ER alpha highlight the MnPN as a potential target for estradiol s sex-specific effects on sleep. Indirect evidence suggests that the VLPO is sensitive to changes in estradiol. Neuronal activation and protein expression of L-PGDS within the VLPO is lower in estradiol-treated females compared to oil-treated (293). Others have shown that estradiol reduces mrna expression for L-PGDS in the VLPO and leptomeninges in mice (330). Estradiol also reduces mrna expression of the adenosine 2A receptor, a receptor critical for sleep induction by the somnogen adenosine, in the POA and VLPO of female mice (330). Together, these data suggest that estradiol alters critical factors involved in the 73

89 induction of sleep in the VLPO but it remains unknown if these are direct or indirect effects of estradiol. Estradiol may act within in the MnPN and VLPO to regulate the homeostatic response to sleep deprivation. Estradiol suppresses the homeostatic drive to sleep following sleep deprivation and strengthens the circadian factors influencing the time to sleep (300,301). The MnPN and subpopulations of VLPO cells are responsive to sleep pressure and thought to mediate sleep homeostasis (88). The presence of ER in the POA, as well as our finding of an estradiol-mediated reduction in sleep drive, strongly implicate a direct role of estradiol in the MnPN and VLPO. We hypothesize that the sleep-active POA is a key site for estradiol-mediated changes in sleep in females. We predict that antagonism of ERs in the MnPN will attenuate the suppression of sleep following estradiol administration. Conversely, we do not anticipate a change in estradiol s suppression of sleep following antagonism of ERs in the VLPO due to lower expression of ER alpha compared to the MnPN. Furthermore, we predict that the signaling cascade initiated by estradiol bindings to ER in the MnPN is not only necessary but also sufficient to suppress sleep. Therefore, we anticipate that direct infusion of estradiol into the MnPN will induce wake and suppress sleep. Lastly, we predict that the production of peripheral somnogens that are involved in the induction of sleep, particularly prostaglandin D2 and adenosine from the leptomeninges, is not influenced by estradiol [for review (342)]. In the following experiments, we micro-infused the ER antagonist, ICI 182,780 (ICI), either into the MnPN or the VLPO. Following infusion, we assessed sleep-wake behavior at baseline and following a systemic injection of EB. Next, we tested whether 74

90 estradiol directly infused into the MnPN is sufficient to recapitulate the systemic effects of estradiol. Finally, we systemically administered ICI to address whether peripheral actions of estradiol, particularly in the leptomeninges, play a role in estradiol s suppression of sleep. Methods Animals All experiments were performed in adult, female Sprague-Dawley rats ( g). Animals were housed under a 12h light, 12h dark cycle with free access to food and water for the duration of the study. All procedures were performed in accordance with the National Institutes of Health guide for care and use of laboratory animals. All experiments were approved by and were in accordance with the guidelines of the University of Maryland Institutional Animal Care and Use Committee. Surgeries All animals were OVX and simultaneously fitting for a TL11M2-F40-EET transmitters (DSI), according to our previously described protocols (see General Methods). For animals used in micro-infusion experiments, either (i) a single guide cannula (C315G, 26-gauge; Plastics One) targeted to the MnPN was implanted at a 9 o angle at the stereotaxic coordinates 0.04mm posterior/ +1.0mm lateral/ 6.5mm ventral relative to bregma or (ii) a bilateral guide cannula (C235G, 26-gauge; Plastics One) targeted to the VLPO was implanted at the stereotaxic coordinates 0.1mm posterior/ 1.0mm lateral/ 7.0mm ventral relative to bregma. The cannula and EEG leads were 75

91 secured together with dental cement. Dummy stylets were used to cover and maintain the guide. The skin along the head was sutured closed around the guide and dummy cap, and the dorsal incision was closed with wound clips. All animals were treated with antibiotic ointment and topical lidocaine as well as carpofen (5 mg/kg) postoperatively and then allowed 7 days to recover before the start of the experiments. Hormone replacement Following a within-animal design, all animals received one injection of sesame oil (baseline) followed by two injections, 24h apart of estradiol benzoate (EB; 5 µg, then 10 µg; Sigma-Aldrich, St. Louis, MO) in sesame oil, as described in the General Methods (Fig. 9). Drugs and Infusion Paradigm Animals were randomly assigned into either the vehicle [VEH; 0.25% dimethyl sulfoxide (DMSO) in sterile water] or ICI (50ng in 0.25% DMSO in sterile water; Sigma- Aldrich) infusion groups, and reversed the following week for a second round of infusions. For targeted infusions to the MnPN, the dummy stylet was removed and replaced with a 33-gauge micro-infusion needle (Plastics One), which extends 0.5mm below the tip of the guide cannula. For targeted infusions to the VLPO, the dummy stylet was removed and replaced with a 33-gauge micro-infusion needle, which extends 2.0mm below the tip of the guide cannula. The needle was connected to a Hamilton 1705 RNR 50ul syringe (Hamilton, Reno, NV) via polyethylene tubing. A BASi Bee pump and Bee Hive controller (Bioanalytical Systems, Inc., West Lafayette, IN) was used to deliver ICI 76

92 or VEH at a rate of 0.1µl/min. Following infusion, the needle remained in place for 5 minutes to ensure diffusion. ICI or VEH was infused 3 times per injection: (i) 6-12h prior to, (ii) 30 minutes prior to and (iii) 12h after injections (Fig. 18A). In a separate cohort of OVX animals, either water-soluble estradiol (3ug cyclodextrin-encapsulated 17β-estradiol in sterile saline, Sigma-Aldrich) or VEH [3ug 2- hydropropyl- β-cyclodextrin (CD) in sterile saline; Santa Cruz Biotechnology, Dallas, TX] was infused into the MnPO at lights on (ZT0) in the absence of systemic estradiol (Fig 18B). Infusions of soluble estradiol or VEH were at a rate of 0.1 µl/min and the needle remained in place for 5 minutes following infusion to ensure diffusion. Figure 18. Injection and infusion paradigm for ER-antagonism studies. (A) ICI infusions in the MnPN and VLPO, (B) Estradiol infusions into the MnPN, and (C) Peripheral ICI study. Arrows indicate timing of infusion, solid lines indicate timing of injection 77

93 In the last cohort of animals, systemic injections of ICI [1mg/kg ICI in 0.01% ethanol (EtOH) in oil] or VEH (1mg/kg 0.01% EtOH in oil) were simultaneously administered with oil/eb injections at lights out (Fig. 18C). Cannula placement At the end of the experiment, animals were overdosed with a ketamine/ acepromazine mix before being transcardially perfused with 0.9% saline + 2% sodium nitrite followed by 4% paraformaldehyde in 0.05M KPBS. The brains were removed and post-fixed overnight in 4% paraformaldehyde. Brains were cryoprotected in 30% sucrose in KPBS, frozen on dry ice, and stored at -80 o C. Each brain was cut on a cryostat along the coronal plane at 30µm thick into 4 series and stored in an ethylene glycol-based storage solution at -20 o C. Sections in each series are separated by 120µm. Sections corresponding to the MnPN and VLPO from one series were mounted on 2% gelatin-coated slides. The slides were processed for cresyl violet (0.1% solution; cresyl violet acetate, Sigma-Aldrich) staining to examine cannula placement. MnPN hits were counted as placement within sections of the brain atlas (343) and VLPO hits were counted as placement within sections For the MnPN cannulations, animals with cannula placement outside of this area were removed from analysis; 3 animals were removed. There was 1 animal whose cannula placement was a miss but she remained in the analysis. This animal was infused with VEH and her behavior was not different from VEH-hits. For the VLPO, one animal was a miss and excluded from the study, while 3 animals were euthanized prior to completion and therefore removed from the study. 78

94 Data analysis The scored epochs were summed over the 12h light phase and 12h dark phase and reported as the total time (in minutes) spent in each state (wake, total sleep, NREMS, and REMS). Within each 12h phase, the percent change induced by estradiol was calculated for each vigilance state Percent change from oil baseline = [(estradiol_time baseline_time) / baseline_time] x 100. Statistical Analysis All data are represented as mean ± SEM. Two-way, repeated measures ANOVAs followed by Bonferroni post-hoc tests were run for each vigilance state to determine if direct MnPN and VLPO infusions significantly altered estradiol s effects on sleep-wake. Since this was a within-animal study, systemic injection (oil vs. estradiol) was the repeated factor and infusion (VEH vs. ICI) was the independent factor. An a prior comparison of interest was between the VEH and ICI infused estradiol days of analysis. We ran an unpaired t-test to compare means on the estradiol day between VEH and ICI infused animals. T-tests were used to compare estradiol and VEH MnPN infusions and two-way, repeated measures ANOVAs followed by Bonferroni post-hoc tests were run for analysis across the phase in 1h bins. Mann-Whitney U nonparametric tests were run to analyze differences between mean percent changes of each vigilance state. Results Direct infusion of ICI into the MnPN attenuates estradiol s suppression of sleep. In all groups, systemic estradiol significantly modulated sleep-wake behavior (Table 3). There was a main effect of estradiol treatment for wake (F 1,12 = 53.48, P < 79

95 Table 3. Effects of direct infusion of ICI into the MnPN on the duration of sleep-wake behaviors in oil and estradiol-treated females VEH (n=6) Dark Phase ICI (n=8) Oil Estradiol Oil Estradiol Wake ± ± 16.0* ± ± 12.3* Sleep ± ± 16.0* ± ± 12.3* NREMS ± ± 14.4* ± ± 10.2* REMS 34.6 ± ± 2.4* 37.1 ± ± 3.1* Light Phase VEH ICI Oil Estradiol Oil Estradiol Wake ± ± ± ± 14.5 Sleep ± ± ± ± 14.3 NREMS ± ± ± ± 16.0 REMS 64.8 ± ± 3.6* 70.8 ± ± 6.0 Data are represented as mean ± SEM. *, P < 0.05 vs. oil ), total sleep (F 1,12 =53.48, P < 0.001), NREMS (F 1,12 =39.93, P < 0.001) and REMS (F 1,12 =57.03, P < 0.001) during the dark phase (Table 3), as well as for REMS (F 1,12 =12.24, P =0.004) in the light phase. Estradiol treatment increased the time spent in wake and decreased sleep, both NREMS and REMS during the dark phase. In this study, pairwise comparisons of VEH infused animals given oil then estradiol revealed that estradiol treatment increased wake duration (t 5 = 2.56, P = 0.05) and decreased total sleep (t 5 = 2.68, P = 0.04) and REMS (t 5 = 2.78, P = 0.04; Table 3). Direct infusion of ICI into the MnPN significantly attenuated these effects during the dark phase (Fig. 19). Animals who received direct infusions of ICI into the MnPN acquire about 47 minutes less wake (t 12 = 2.376, P = 0.04) than VEH and about 37 more minutes of NREMS (t 12 = 2.158, P = 0.05) and 10 more minutes of REMS (t 12 = 2.518, P = 0.03; Fig.19A) during the dark phase. The percent change in wakefulness induced by estradiol was not significantly 80

96 different between VEH and ICI infusion groups (U = 16.0; P = 0.35; Fig.19B). The percent changes in total sleep (U = 7.0; P = 0.03), NREMS (U = 8.0; P = 0.04) and REMS (U = 2.0; P = 0.003) induced by estradiol were significantly attenuated by ICI. The percent changes in wake, total sleep, NREMS and REMS were not affected by ICI during the light phase (data not shown). Direct infusion of estradiol into the MnPN increases wake and suppresses sleep. Estradiol infusion occurring at the beginning of the light phase, significantly increased wake (t 12 = 2.20; P = 0.05; Fig. 20A) and decreased total sleep (t 12 = 2.19; P = 0.05; Fig. 20B) over 6h after infusion. The first 6h and the last 6h of the dark phase were unaffected by direct estradiol infusion into the MnPN. Analysis of wake during the dark period in 1h bins revealed a significant main effect of the infusion (F 1,60 = 4.86, P = 0.05) and an interaction (F 5, 60 = 2.78, P = 0.03). Across the 6h period, wakefulness was elevated following estradiol infusion but did not reach significance until ZT 11 (t = 3.25, P < 0.05; Fig. 21A). For total sleep, there was a significant main effect of the infusion (F 1,60 = 4.82, P = 0.05) and an interaction (F 5, 60 = 2.79, P = 0.03). Total sleep was significantly decreased during ZT 11 (t = 3.26, P < 0.05). There was a main effect of the time (F 5,60 = 2.61, P = 0.03) and an interaction (F 5, 60 = 3.32, P = 0.03) for REMS. Across the 6h period, REMS was decreased following estradiol infusion but did not reach significance until ZT 11 (t = 3.76, P < 0.05; Fig. 19D). There were no significant effects found for NREMS. Wake, total sleep, NREMS, and REMS were not different during across the dark phase (Fig. 21E-H). 81

97 Figure 19. Effects of direct infusion of ICI into the MnPN on estradiol-mediated changes in sleep. (A) Antagonism of ER in the MnPN significantly attenuates the effects of estradiol on wake, total sleep, NREMS, and REMS. (B) The change in sleep, both NREMS and REMS, is significantly attenuated by direct infusion of ICI in to the MnPN, by about 50%. *, P <0.05 vs. VEH. 82

98 Figure 20. Effects of direct infusion of estradiol into the MnPN on sleep and wakefulness. (A) Wakefulness was increased 6h following a local infusion of estradiol into the MnPN. (B) Total sleep is reduced 6h following infusion of estradiol into the MnPN. (C-D) NREMS and REMS were not significantly decreased. (A-D) Sleep-wake behavior was not affected during the dark phase. 83

99 Figure 21. Effects of direct infusion of estradiol into the MnPN on sleep and wakefulness across the light and dark phase. Local infusion of estradiol at ZT0 significantly increased (A) wake and suppressed (B) sleep, specifically (D) REMS during ZT 11. (E-H) Sleep-wake behavior across the dark phase was unaffected. *, P < 0.05 vs. Vehicle. 84

100 Direct infusion of ICI into the VLPO may attenuate estradiol s suppression of sleep. Preliminary findings suggest that ICI infusion into the VLPO does not attenuate estradiol s effects on wake and sleep (Fig. 22; Table 4). However, the variability in REMS data makes it difficult to interpret any potential effect of ICI. This study is ongoing in the lab and more animals will be added. We calculated the effect size of this preliminary cohort for total sleep (d=0.52) and REMS (d=0.24). These effect sizes are lower than what we calculated for our preliminary cohort when we began the MnPN infusion experiments (total sleep d=0.67 and REMS d=1.36). Estradiol does not affect peripheral somnogens to suppress sleep. Systemic injections of ICI did not affect estradiol s ability to suppress sleep or increase wake (P > 0.05). Percent change data for wake, total sleep, NREMS and REMS are presented in Figure 23. Table 4. Effects of direct infusion of ICI into the VLPO on sleep-wake behaviors in oil and estradiol-treated females. VEH (n=4) Dark Phase ICI (n=3) Oil Estradiol Oil Estradiol Wake ± ± ± ± 37.9 Sleep ± ± ± ± 37.9 NREMS ± ± ± ± 33.8 REMS 37.0 ± ± ± ± 9.6 Data are represented as mean ± SEM. 85

101 Figure 22. Effects of direct infusion of ICI into the VLPO on estradiol-mediated changes in sleep and wake. (A-B) ICI did not attenuate estradiol s effects on sleep-wake. n=4 for VEH, n=3 for ICI. 86

102 Figure 23. Effects of a systemic injection of ICI on sleep and wake following estradiol treatment. In both VEH and ICI treated animals, estradiol increased wake and suppressed sleep, both NREMS and REMS. Peripheral ICI did not attenuate the effects of estradiol on sleep and wake. 87

103 Discussion Here, we show that the sleep-active POA is a direct target for estradiol s suppression of sleep. The MnPN is a necessary site of estradiol action for its full suppression of sleep. Antagonism of ERs in the MnPN attenuated the estradiol-mediated suppression of sleep, by ~50%. Preliminary data from ongoing studies in the lab indicate that the VLPO may not be critical for estradiol s effects as the MnPN. We also found that peripheral effects of estradiol did not suppress sleep. Together, our findings begin to elucidate key sites for estradiol within the sleep-wake circuitry. In years past, much research has focused on understanding the role of estradiol in the induction of arousal and it was unclear if estradiol directly affected sleep-promoting nuclei [for review (344)]. Both the MnPN and VLPO have been extensively studied for their role in sleep initiation and maintenance [for review (94,95,102)]. Both nuclei express ER, with the MnPN having markedly higher expression compared to the VLPO. Here, we found that ER-activity in the MnPN was critical for estradiol s suppression of sleep. Direct infusion of ICI into the MnPN attenuated the estradiol-mediated suppression of sleep, both NREMS and REMS, suggesting that ER-action in the MnPN is necessary for estradiol s full sleep suppressive effect. These effects were specific to the dark phase. OVX females treated with estradiol had increased sleep and decreased wake compared to oil treatment during the light phase. These findings support the hypothesis that estradiol strengthens circadian factors regulating sleep; however, ER antagonism did not block this effect. Micro-infusion of estradiol into the MnPN significantly modulated sleep and wakefulness. Estradiol increased wake and suppressed sleep compared to VEH infusion. 88

104 Here, the effects of a local estradiol infusion were less in magnitude and shorter in duration. Estradiol increased wake and suppresses sleep across the last 6h of the light phase, but did not have an effect on sleep-wake behavior during the following dark period. Furthermore, the effects were greatest during the last hour of the light phase, suggesting that transition periods may be most sensitive to estradiol. Based on our findings, we predict that estradiol directly suppresses sleeppromotion. Antagonism of ER partially blocked the estradiol-mediated suppression of both NREMS and REMS and activation of ER via direct infusion of estradiol suppressed total sleep for a short period of time. Estradiol may suppress sleep by either reducing excitatory input or by hyperpolarizing resting membrane potential of MnPN neurons. Direct modulation of the MnPN neuronal activity affects the activity of the orexinergic neurons in the PF/LH, such that activation of the MnPN inhibits the PF/LH (110,345,346). Therefore, one likely downstream effect of estradiol decreasing MnPN activation would be a release of the inhibitory tone on orexinergic neurons. The direct role of estradiol in the VLPO remains unclear. Previously, we reported that exogenous estradiol reduces the activation of VLPO neurons (293). Here, our preliminary data suggest the VLPO is not likely a critical target for estradiol s suppression of sleep. Our effect size calculations indicate that any effects of ICI in the VLPO will most likely be in total sleep but not REMS and may not be as robust as in the MnPN. Peripherally derived somnogens can also mediate VLPO activity [for review (342)]. Estradiol reduces L-PGDS transcript levels in the leptomeninges of the POA (347), potentially representing a peripheral effect of estradiol. However, peripheral ICI at a dose that does not to cross the blood brain barrier did not block estradiol s effects on 89

105 sleep and wakefulness, but did block the increase in uterine horn weight (348). These data suggest that peripherally derived somnogens are unaffected by estradiol. Ongoing studies in the lab are exploring the effect of peripheral ICI on the expression of L-PGDS in the leptomeninges to determine if ICI was able to gain access to the meningeal layer. Together, these data highlight a key area within the sleep-circuitry that estradiol targets to suppress sleep, but it is clear that estradiol has other targets as well. Local infusion of ICI into the MnPN only attenuated estradiol s effects by about 50%. It is likely that estradiol acts more globally, like in the arousal nuclei of the brainstem, to enhance arousal. The identification of the MnPN as a direct site of estradiol action now allows for more mechanistic research in these regions to determine how estradiol is suppressing sleep in females. 90

106 Chapter 6: Orexin is not the mediator of estradiol s effects on sleep Introduction The sleep-active nuclei in the POA have dense projections to many nuclei involved in the onset and maintenance of arousal. The MnPN contains ERs and antagonism of these receptors attenuated the estradiol-mediated suppression of sleep (previously described in Chapter 5). What remains unclear is what downstream signaling is changed by direct estradiol action in the MnPN. Orexin, an arousal neuropeptide, is an ideal candidate for mediating estradiol s effects. Orexin is involved in arousal, consolidates or stabilizes vigilance states, and it is a potent inhibitor of REMS. Orexin receptors (orexin receptor 1 and 2) are G-protein receptors coupled to Gq, which have excitatory effects on target cells [for review (349)]. Both the MnPN and VLPO send inhibitory projections to the orexinergic neurons in the PF/LH (96-99,109). In fact, there is evidence that MnPN and PF/LH neurons reciprocally modulate activity, such that activation of the MnPN reduces arousal related firing and increases sleep-active activity in the PF/LH (110). Orexinergic neurons send projections to arousal nuclei in the brainstem, like the locus coeruleus (LC) and dorsal raphe nucleus (DRN), which are also targets of the MnPN and VLPO (350). Modulation of MnPN activation by estradiol may lead to changes in orexin signaling inducing arousal. The orexinergic system is influenced by estradiol. Females treated with estradiol have greater expression of activated orexinergic neurons compared to oil-treated females (329). Fluctuation in endogenous gonadal steroids across the estrous cycle, as well as treatment with exogenous estradiol in OVX females, increases expression of orexin (351) 91

107 and its receptors ( ). Additionally, prepro-orexin mrna expression in the hypothalamus is higher in females compared to males (355). We hypothesize that orexin is the mediator of estradiol s suppressive effects on sleep. To test this, we obtained dual orexin receptor antagonists (DORA-12), which blocks the activation of orexin receptors 1 and 2. DORA-12 promotes both NREMS and REMS in rats (356). We predict that if orexin were the mediator of estradiol s suppressive effects on sleep then antagonism of orexin receptors would attenuate the estradiol-mediated suppression of sleep. Additionally, due to the sex difference in preproorexin mrna and steroidal modulation of receptor expression, we predict that sleeppromotion by DORA-12 will be sex and estradiol dependent. We also quantified the number of Fos-ir cells, orexin-ir cells, and Fos+ orexinergic-ir cells in the PF/LH when ICI attenuates estradiol s effects on sleep. We predict that if estradiol increases orexin activation to suppress sleep, then ICI infusion into the MnPN will attenuate the increase in colabeled cells in the PF/LH following estradiol treatment. Methods Experiment 1: Effects of DORA-12 Animals and Drug Paradigm Gonadectomized female ( g) and gonadally intact male ( g) rats were fitted with EEG/EMG transmitters (DSI). All animals were treated with antibiotic ointment and topical lidocaine as well as carpofen (5mg/kg) post-operatively, and then allowed 7 days to recover before the start of the experiments. 92

108 All female rats were treated with oil and EB following our established hormone replacement paradigm, while males just received oil injections as a control. Animals were randomly assigned to either DORA-12 or vehicle treatment groups. Following a wash-out period of 5 days, all animals were reassigned to the other treatment group. DORA-12 was suspended in 20% D-α-tocopheryl polyethylene glycol 1000 succinate (Vitamin E TPGS) at a concentration of 10mg/ml. First, we ran a dose response (n=2 per group) to determine the concentration to use in females. DORA-12 (3mg/kg, 10mg/kg, or 30mg/kg) or vehicle (VEH; equal volume of Vitamin E TPGS) was given orally prior to lights out (ZT12). For the experiment, DORA-12 (30mg/kg) or VEH was given orally prior to lights out (ZT12) each day of injections (Fig. 24). Figure 24. Gavage and hormone replacement paradigm. Arrows indicate injection times, lines indicate administration of DORA-12. Sleep Analysis The total duration (in minutes) for each vigilance state was analyzed for the dark phase for each scoring period. To compare the magnitude of change induced by either estradiol replacement or DORA-12, the percent change was calculated [((treatedbaseline)/baseline) *100]. 93

109 To test whether orexin is the mediator of estradiol s suppressive effects on sleep, we first compared the total wake, sleep time, NREMS and REMS between oil and estradiol days in VEH and DORA-12 treated females using a two-way repeated measured ANOVA. The repeated factor was treatment (oil vs estradiol) and independent factor was drug treatment (VEH or DORA-12). Next, we compared the sleep-promoting effect of DORA-12 in males, oil-treated females, and estradiol-treated females to test whether sleep promotion by DORA-12 is sex and estradiol dependent. Here, we looked at the amount of wake across the 12h dark phase in 3h bins. Two-way repeated measures ANOVAs were run for each group. The repeated factor was time and independent factor was drug treatment (VEH or DORA-12). A student s t-test was used to compare the total time in wake for VEH vs DORA-12 treated animals in each group. Lastly, comparisons for the DORA-12 treatment alone were done between all groups. Two-way repeated measures ANOVA was used compare wake time following DORA-12 in males, oil-treated females, and estradiol-treated females across the dark phase in 3h bins. A one-way ANOVA was used to compare total wake time in the dark phase following DORA-12 in males, oil-treated females, and estradiol-treated females. A Mann-Whitney U non-parametric t-test was used to compare the percent change in each wake during 3h intervals across the dark phase in females. Experiment 2: Effects of ICI infusion into the MnPN on orexinergic activation A cohort of female Sprague-Dawley rats ( g) was simultaneously implanted with EEG/EMG telemetry transmitters (DSI) and a single guide cannula targeted to the MnPN. Following the infusion paradigm described in Chapter 5, all 94

110 females were overdosed with a ketamine/acepromazine mix before being transcardially perfused with 0.9% sodium chloride + 2% sodium nitrite solution followed by 4% paraformaldehyde in 0.05M KPBS 5h into the dark phase (ZT 17). The brains were removed, postfixed overnight, cryoprotected and stored -80 o C. Briefly, each brain was cut on a cryostat along the coronal plane at 30µm thick into 4 series and stored in an ethylene glycol-based storage solution at -20 o C until processed. Sections in each series are separated by 120µm. Cohorts containing animals from all treatment groups [(i) VEH-Oil, (ii) VEH-EB, (iii) ICI-Oil, (iv) ICI-EB] were immunocytochemically processed in the same tray. Sections were first processed for Fosir for 48h at 4 o C with rabbit IgG anti-c-fos (Ab-5) antibodies corresponding to amino acids 4-17 of human c-fos (1:50,000; Calbiochem PC38) and visualized with nickel sulfate-enhanced DAB in TBS solution containing 0.005% H 2 O 2. The sections were rinsed overnight and then processed for orexin-ir using goat polyclonal IgG anti-orexin A antibodies (1:20,000; Santa Cruz SC-8070) for 48h at 4 o C and visualized with DAB in TBS solution containing 0.005% H 2 O 2. Sections were mounted serially on 2% gelatincoated glass slides and coverslipped. We systematically counted the number of Fos-ir, orexin-ir, and colabeld cells in the PF/LH with the aid of the Neurolucida software as previously described (293). Sections analyzed were from two-in-four series (adjacent sections were separated by 60 µm). Three brain sections corresponding to Plates of the Paxinos and Watson rat brain atlas were used in the analysis (90). We used an 800µm x 1200µm contour; with the ventral edge was center on the ventral edge of the fornix. Two contours per bilateral 95

111 section of the PF/LH were counted and a total number of Fos-ir, orexin-ir, and co-labeled cells were averaged between the two series. Results Dose response for DORA-12 in females. In females, DORA-12 (30mg/kg) significantly suppressed wake (F 3,4 = 13.1, P = 0.016; Fig 25A) and increased sleep (F 3,4 = 12.19; P = 0.018; Fig. 25B) compared to VEH. The lower doses of DORA-12 (3mg/kg and 10mg/kg) were not efficient at reducing 12h total time in wake and promoting sleep compared to VEH (Fig. 25). Orexin is not the mediator of estradiol s suppressive effects on sleep DORA-12 significantly suppressed wake and increased sleep, both NREMS and REMS in oil and estradiol-treated females (Fig. 26). There was a significant main effect of DORA-12 treatment for wake (F 1,13 = 111.4, P < 0.001), total sleep (F 1,13 = 109.1, P < 0.001), NREMS (F 1,13 = 80.6, P < 0.001) and REMS (F 1,13 = 49.5, P < 0.001). There was also a significant main effect of hormone treatment for wake (F 1,13 = 32.4, p < 0.001), total sleep (F 1,13 = 33.2, P < 0.001), NREMS (F 1,13 = 13.4, P = 0.003) and REMS (F 1,13 = 7.6, P < 0.001), such that estradiol significantly increased wake (Fig. 26A) and suppressed sleep (Fig. 26B), specifically REMS (Fig. 26D) in both VEH and DORA-12 treated females. Estradiol did not, however, suppress NREMS in DORA-12 treated females (t 13 = 1.89, P > 0.05; Fig. 26C). 96

112 Figure 25. DORA-12 dose response. (A) DORA-12 significantly suppressed 12h wake duration in females treated with 30mg/kg, but not 3mg/kg or 10mg/kg. (B) DORA-12 significantly increased total sleep duration during the 12h dark phase following administration in females treated with 30mg.kg, but not 3mg/kg or 10mg/kg. n=2 per group. *, P <

113 Figure 26. Effects of DORA-12 in oil and estradiol-treated females. DORA-12 suppressed wake and promoted sleep, both NREMS and REMS in all animals. Estradiol was able to (A) increase wake and (B) suppress sleep. (C) NREMS was unaffected by estradiol in DORA-12 treated, but (D) REMS was significantly reduced. *, P < 0.05 from oil. 98

114 The sleep-promoting effect of DORA-12 is sex dependent in rats In males, there was a significant main effect of time (F 3,12 = 7.50, p = 0.004; Fig. 27A). DORA-12 treatment significantly suppressed dark-phase wake for the 3h period following administration in males (t=2.92, P < 0.05). It did not significantly affect the 12h total wake in the dark phase (t 4 =1.51, P > 0.05). In oil-treated females, there was a significant main effect of both drug treatment (F 1,13 = 76.88, P < 0.001) and time (F 3,39 = 21.37, P < 0.001) and there was a significant interaction (F 3,39 = 3.11, P = 0.04; Fig. 27B). DORA-12 significantly suppressed wake during the first 3h (t=5.65, P < 0.001) following administration and the last 6h of the dark phase (t=5.22, p < 0.001; t = 5.13, P < 0.001). DORA-12 significantly suppressed total wake time in the dark phase by ~ 40% (t 13 = 8.77, P < 0.001). In estradiol-treated females (Fig. 27C), there was a significant main effect of drug treatment (F 1,13 = 66.92, P < 0.001) and trend of time (F 3,39 = 2.47, P = 0.08). DORA-12 significantly suppressed wake across the entire dark phase (each 3h bin: t = 5.43, p < 0.001; t = 4.39, P < 0.001; t = 5.05, P < 0.001; t = 4.43, P < 0.001). DORA-12 significantly suppressed total wake time in the dark phase by ~ 37% (t 13 = 8.18, P < 0.001). Using a two-way RM ANOVA analyzing DORA-12 effects in all groups, there was a significant main effect of both group (F 2,18 = 14.50, P = 0.002; Fig. 28A) and time (F 3,54 =18.58, P < 0.001) and there was a significant interaction (F 6,54 = 2.48, P = 0.03). The magnitude of DORA-12 s wake suppressing effect was significantly attenuated in males compared to oil-treated females across the entire dark phase (Fig. 28B) and during ZT compared to estradiol-females (t = 4.41, P < 0.05; Fig. 28B). The magnitude of DORA-12 s effect was significantly attenuated in males compared to estradiol-treated females during ZT 21-0 (t = 3.71, P < 0.05; Fig. 28B). 99

115 Figure 27. Effects of DORA-12 across the 12h dark period. (A) In males, DORA-12 suppressed wake 3h post-administration, but it did not affect the 12h wake total. (B) In oil-treated females, DORA-12 significantly suppressed wake during most of the dark phase, reducing the 12h wake total by ~40%. (C) In estradiol-treated females, DORA-12 significantly suppressed wake during across the dark phase, reducing the 12h wake total by ~37%. *, P < 0.05 from vehicle. 100

116 Figure 28. The sleep-promoting effect of DORA-12 is sex dependent. (A). Following DORA-12 administration, wake was significantly more suppressed in oil-treated females compared to males across the dark phase. (B) 12h totals for wake in the dark phase were significantly different between all groups. a, P < 0.05 from oil; b, P < 0.05 from male; c, P < 0.05 from estradiol. 101

117 Estradiol does not influence the efficacy of DORA-12 Since estradiol increases wake, we calculated the percent change in wake following DORA-12 in both oil- and estradiol-treated females to determine if the differences listed above are due to a change in baseline wake or if estradiol influences the efficacy of DORA-12. The percent change in wake following DORA-12 is not significantly different across the 12h dark phase between oil and estradiol-treated females (Fig. 29). Effects of ICI infusion into the MnPN on orexinergic neuronal activation Preliminary data point to potential changes in Fos-ir (Fig. 30A) and orexin-ir (Fig. 30B) in the PF/LH following ICI infusion in the MnPN. However, due to a small sample size, statistical power is lacking to draw definitive conclusions. The current data suggest that ICI infusion into the MnPN does not affect the activational state of orexinergic neurons in the PF/LH (Fig. 30C). Lastly, these findings indicate that estradiol treatment does not increase orexinergic activation compared to oil treatment. Discussion In this set of experiments, we tested the role of orexin in estradiol s suppression of sleep. First, we found that antagonism of both orexin receptor 1 and 2 did not completely block the estradiol-mediated suppression of sleep, especially REMS. We found that wake suppression by DORA-12 was sex dependent, such that the sleeppromoting effects of DORA-12 persisted longer in females compared to males. Furthermore, we found that the activational state of orexinergic neurons was unaffected 102

118 Figure 29. Effects of hormonal status on DORA-12 efficacy. Wake suppression was similar in oil- and estradiol-treated females. 103

119 Figure 30. Effects of ICI infusion in the MnPN on orexinergic activation. (A) Fos-ir, (B) orexinir, and (C) co-labeled cells may be elevated in ICI-oil animals compared to all other groups. 104

120 by direct ICI infusion into the MnPN or by estradiol treatment. Together, these findings indicate that orexin is not increased following estradiol s inactivation of the MnPN and does not mediate estradiol s suppression of sleep and induction of arousal. Our finding that antagonism of ERs in the MnPN attenuated the estradiolmediated suppression of sleep (see Chapter 5) strongly implicates the orexinergic neurons as potential downstream targets of estradiol action in the MnPN. It is well known that orexin plays a key role in arousal, particularly in the stabilization of vigilance. Orexinergic neurons receive dense projections from the MnPN and this innervation has been shown to directly regulate orexinergic activity (110). Our lab has preliminary data indicating that estradiol increased orexin activation, which was similar to previously reported findings (329). These findings further supported our hypothesis that orexin is a key downstream target of estradiol s action in the MnPN. In our current study, we did not replicate our finding that estradiol increases the activation of orexinergic neurons. Unlike our earlier findings and those by Deurveilher and colleagues (329), these data were collected 5h into the dark phase, whereas the others were collected during the light phase. Since this was a small sample, more animals need to be added before we can draw any definitive conclusions, but there may be circadian differences in hormonal modulation of orexinergic activation. Since estradiol did not increase orexinergic activation, it was difficult to assess the effects of ICI infusion in the MnPN on orexinergic activation. ICI-Oil animals seemed to have the highest level of orexinergic action compared to all groups. It is possible that there is local synthesis of estradiol in the MnPN, which has basal effects on MnPN neuronal activity, or that ICI has additional non-antagonistic effects. 105

121 A direct test of our hypothesis was assessing the effects of estradiol on sleep following administration of dual orexin receptor antagonists. DORA-12 significantly suppressed wake and promoted sleep in both oil- and estradiol-treated females. When we compared the effects of estradiol treatment to oil in animals receiving the DORA-12, we found that estradiol was still able to significantly induce arousal at the expense of sleep. This suppression of sleep, however, was mainly due to estradiol suppressing REMS, as NREMS suppression was attenuated in DORA-12 females. These data strongly suggest that orexin is not the key mediator of estradiol s suppression of sleep. It is likely that orexinergic signaling is increased by estradiol since DORA-12 was able to block the suppression of NREMS; however, other neurotransmitters are probably involved as well, leading to the more robust suppression of sleep. There are many other potential targets underlying estradiol s suppression of sleep. One such target may be histamine. The histaminergic neurons in the TMN are under direct inhibitory control of VLPO neurons. Previously, we found that estradiol causes a decrease in Fos expression in the VLPO and a concomitant increase in Fos expression in the TMN (293). It is possible that estradiol action in the MnPN feeds forward to the VLPO, which in turn releases the break on histamine, promoting wake. This suggested mechanism supports the flip-flop model proposed by Saper et al. to switch between sleep and wake states (102). In the current study, we found that the sleep-promoting effect of DORA-12 was sex dependent in rats. In males, DORA-12 suppressed wake for ~3h post-administration and does not affect overall time awake during the dark period. In females, the sleeppromoting effect of DORA-12 was more robust, significantly suppressing wake by ~40% 106

122 in the dark phase. DORA-12 similarly suppressed wake in both oil and estradiol-treated animals. Baseline levels of wake are higher in estradiol-treated females so the total amount of wake was significantly higher in estradiol vs. oil DORA-12 treated females. Sex differences in hepatic drug metabolism have been well defined in humans and rodents [for review (357,358)]. Differences in the hypnotic effect of barbiturates were the first reports of sex differences in drug metabolism, such that males slept less than females following drug administration (359,360). To determine if our findings are simply due to a sex difference in the metabolism of DORA-12, we are running a pharmacokinetic study to determine plasma levels of DORA-12 1, 2, 4 and 6h following administration in males and oil- and estradiol-treated females. Sex differences in the organization and/or sensitivity of the sleep circuitry can affect the modulation of sleep by various drugs. We have previously described (Chapter 4) that there was a sex difference in the sleep circuitry, particularly in the VLPO. It remains unclear if there are sex differences in the MnPN. In humans, olanzapine has sexspecific effects on sleep that are independent of drug metabolism (361). Women show an increase in SWS following olanzapine, while men show a decrease in SWS (361). It is possible then, that sex differences in circuitry can underlie the differences we see here with sleep-promotion by DORA-12. Sex differences and hormonal modulation of the orexinergic system may also contribute to this difference ( ,355). To directly test if this difference is due to sexually dimorphic circuitry, we plan to infuse DORA-12 i.c.v to bypass hepatic metabolism and see if the sex difference in sleep-promotion duration can be recapitulated. 107

123 Overall, these findings suggest that orexin is not the key mediator of estradiol s effects on sleep. While orexin seemed like the ideal candidate, antagonism of orexin receptors 1 and 2 was not able to block the estradiol-mediated arousal and suppression of REMS. Additionally, preliminary data indicate that antagonizing ERs in the MnPN has little to no effect on orexinergic activation. Together, these data suggest that other neurotransmitter systems are involved in estradiol s suppression of sleep. 108

124 Chapter 7: In utero exposure to valproic acid changes sleep in juvenile rats: a model for sleep disturbances in autism Introduction Autism spectrum disorders (ASD) are an array of developmental disorders, primarily disrupting social interactions, communication, and behavioral patterns. The prevalence of ASD is male-biased, with the averaged male-to-female ratio ~4:1 [for review (362,363)]. Clinical observations have also indicated a high prevalence of sleep disturbances in children with ASD [for review ( )]. Recent clinical studies report that the prevalence of sleep problems in children with ASD is roughly 44% to 83% of diagnosed cases ( ). These sleep dysfunctions typically manifest as difficulties initiating and maintaining sleep, sleep fragmentation, insomnia, and parasomnias.(371, ). Quality sleep is imperative for the maintenance of good health and sleep loss can lead to or exacerbate existing behavioral problems associated with ASD (377,384) [for review (72,385)]. Alterations in the sleep architecture have been linked with cognitive and behavioral deficits as well as stress and anxiety [for review (72,385,386)]. Since children diagnosed with ASD share similar symptoms to those experiencing sleep loss, it is plausible to suspect that sleep disturbances in ASD may contribute to its symptomology. However, it appears that sleep problems in ASD are not influenced by the severity of cognitive deficits or ASD subtypes (379,387). Although these sleep disruptions are characterized, their cause remains unknown and relatively unexplored. It is also unclear if there is a sex difference in sleep behavior in ASD. The rodent s neurocircuitry and neurochemistry of sleep share similarities with humans, suggesting that rats would be a good model system for these basic investigations 109

125 (388). To date, the current animal models of ASD have been underutilized in the investigation of ASD associated sleep disturbances. Here, we present data supporting the use of a rodent model in elucidating the etiology of sleep disruptions in ASD. One animal model of ASD is prenatal exposure to valproic acid (VPA) or the salt, sodium valproate. VPA is an antiepileptic drug used to treat seizure disorders. While VPA can prevent the induction of seizures during pregnancies, it has been classified as a teratogen and linked to a clinical phenotype known as fetal valproate syndrome (FVS), which includes various types of craniofacial malformations and developmental and cognitive delays (389) [for review (390,391)]. Many studies have also shown that FVS is associated with autistic behaviors and it is now thought that VPA exposure during a critical period of development significantly increases the risk of ASD (389,392,393). In a rodent model, animals exposed to VPA prenatally show similar alterations in brain architecture and sex-specific behavioral deficits seen in children with ASD ( ). Therefore, it is thought to be a valid animal model for studying ASD. To our knowledge, only one study has indirectly evaluated sleep/wake behavior in an animal model of ASD using locomotor and feeding behaviors (403). With the VPA model, Tsujino and colleagues have shown abnormal circadian rhythms in their animals, marked by increased locomotor activity and feeding behaviors during their sleep phase (403). Additionally, using microdialysis, they found that basal levels of serotonin (5-HT) are higher in VPA-exposed animals and that there is a caudal shift of 5-HT positive neurons in the dorsal raphe nucleus (403). Interestingly, the dorsal raphe supplies the serotinergic input involved in the ascending arousal system of the sleep/wake circuitry. 110

126 In this study, we investigated sleep/wake behaviors, including the sleep patterning, and electroencephalogram (EEG) characteristics related to the individual vigilance states in juvenile rats. The majority of studies investigating sleep in ASD report findings from children and adolescents with ASD; therefore, we recorded from juvenile/prepubertal rats to better model brain maturation and sleep in children and young adults. We were unable to look for sex differences in sleep in this model as we originally intended because we could not generate enough female VPA-exposed animals. Since sleep/wake cycles are controlled by a complex system of reciprocal connections between sleep-promoting nuclei and arousal centers, we investigated whether VPA exposure alters some aspect of the sleep-wake neurocircuitry. Specifically, we investigated whether VPA alters the GABAergic system, which plays a significant role in sleep onset and maintenance [for review (95)], through quantification of the expression of glutamic acid decarboxylase (GAD), the rate-limiting enzyme in gamma-aminobutyric acid (GABA) synthesis. Here, we measured GAD expression in the cortex and basal forebrain (BF), which is involved in sleep/wake and has been shown to regulate activity of cortical neurons (404,405). Methods Animals All experimental procedures were performed in accordance with the NIH guide for care and use of laboratory animals. All experiments were approved by and were in accordance with the guidelines of the University of Maryland Institutional Animal Care and Use Committee. 111

127 Female Sprague-Dawley rats were mated overnight (12 hours) and the next day was designated day one of gestation. Pregnancy was determined by percent weight gain by gestational day 12. Valproic acid sodium salt (purchased from Sigma) was dissolved in 0.9% sterile saline, ph 7.4. Pregnant dams received a single intraperitoneal injection of 400 mg/kg VPA (VPA-exposed) or sterile saline (SAL-exposed) on gestational day Each female was housed individually and raised her own litter. For this study, 2 SALexposed and 2 VPA-exposed litters were used. Pups were weaned on postnatal day 21. The following experiments were performed in SAL- (n=5) and VPA-exposed (n=6) juvenile male and female rats (~70-80g). Tail malformations that manifest as a distal bend or a kink have been reported as a marker of VPA toxicity during development (390,406,407). Only VPA-exposed rats with tail malformations were used in this study (see below). Only 1 VPA-exposed female had a tail malformation. Therefore, we were unable to look at sex differences in sleep in this model. All animals were housed with a littermate under a 12:12 hour, reverse light:dark cycle with free access to food and water for the duration of the experiment. Surgery A bi-potential-lead transmitter (TL11M2-F20-EET, Data Sciences International, Minnesota, USA) was implanted subcutaneously through a dorsal incision of the abdominal region. Two burr holes (0.5mm diameter) were drilled asymmetrically and two dental screws (Plastics One, Roanoke, VA, USA) were implanted into the skull at +2.0mm AP/+1.5mm ML and -6.0mm AP/-1.5 ML from bregma. Animals were treated 112

128 with antibiotic ointment and topical lidocaine, as well as 0.1cc buprenorphine, postoperatively, and allowed 7 days to recover before being reunited with their non- experimental cage mate. Data acquisition and analysis All recordings took place in a designated room shielded from noise or other background disturbances. EEG and EMG waveform data for post-natal days were collected using the Dataquest ART 4.0 software (DSI, Minnesota, USA) set to a continuous sampling mode. Sleep is typically measured as a function of EEG and electromyogram (EMG) activity. For each of the animals, the 24h (12:12 light:dark cycle) period on post-natal day 32 was scored using the NeuroScore software (DSI, Minnesota, USA). EEG EMG waveforms were scored according to visual inspection of 5s epochs into wake (low-amplitude, high-frequency EEG combined with high- amplitude EMG), NREM (high-amplitude, low-frequency EEG combined with low-emg tone) or REM (low-amplitude, high-frequency EEG combined with muscle atonia and occasional muscle twitches). Transitions were scored when 4 or more epochs of the new state were noted within another stage. This rule was followed unless there was a period punctuated by a new state, which then persisted; this was considered a transition period so these epochs were scored as the new state. The scored epochs were summed over the 12h light phase, 12h dark phase, and 24h period and reported as the total time (in minutes) spent in each state. Within each 12h phase, the average bout duration (in seconds) and the number of each bout and transition were also determined for each vigilance state. 113

129 The DSI module for periodogram powerbands in NeuroScore was used to generate the mean power of frequencies present in the three vigilance states (Wake, NREM sleep and REM sleep) across the light and dark. The periodgrams estimate the power for each defined frequency band (delta: Hz, theta: 4-8 Hz, alpha: 8-12 Hz, sigma: Hz, beta: 16-24, and gamma: Hz) on the Fast Fourier Transform (FFT) designed for continuous data. The relative power band value (percentage; desired band over total power in the signal) was determined in 1h bins and then averaged across each 12h phase for each vigilance state. Western Blotting Following the recording period, on postnatal day 40, brains were collected and frozen on dry ice. Sections (180µm) were cut on a cryostat and micropunches were collected from the somatosensory cortex and BF of each animal, as well as three extra age-matched SAL-controls (total SAL n=7). The tissue was then homogenized via sonication in a cell lysis buffer containing a protease inhibitor cocktail. The protein concentration for each sample was determined using a bicinchoninic acid (BCA) assay kit (Pierce, Rockford, IL). Protein (1 µg) of each sample (SAL=7, VPA=6) was loaded into a 10% Tris-glycine SDS-PAGE gel (Invitrogen, Carlsbad, CA), and then the electrophoresed proteins were blotted onto a polyvinyl difluoride membrane (Invitrogen). The membrane was washed in 20 mm Tris-buffered saline solution with 0.05% Tween 20 (T-TBS). The membrane was blocked overnight in 5% powdered milk at 4 o C and incubated the next day in the anti-gad65/67 primary antibody solution (1:30,000 in T- TBS; Millipore AB1511) for 2h at room temperature. Following the primary antibody 114

130 incubation, the membrane was washed three times in T-TBS, and incubated for 1h at room temperature in a anti-rabbit IgG, HRP-linked secondary antibody solution (1:2000 in T-TBS; Cell Signaling Technologies #7074). The Phototype-HRP chemiluminescent system (Cell Signaling Technology, Danvers, MA) was used for detection of the protein recognized by the antisera. To correct for errors in sample loading, the membrane was also probed with an antibody to the housekeeping gene glyceraldehyde- 3-phosphate dehydrogenase (GAPDH; 1:1,000,000; Millipore MAB374) as previously described ( ). Membrane was exposed to Hyperfilm-ECL (Kodak, Rochester, NY) for varying exposure times. The films were then scanned into a computer at 1200 dpi and analyzed using ImageJ64 software ( The optical densities (o.d.) were measured for each individual band and normalized to the o.d. of the GAPDH bands. Statistical Analysis Statistical differences between SAL- and VPA-exposed animals were determined with Student s t-tests for each sleep parameter studied. Differences in the power spectrum densities were determined with a Mann-Whitney U for non-parametric data. A two-way repeated measures ANOVA was used to analyze wake gamma power across the 24h period. Two-way ANOVAs were used to analyze NREM sleep and REM sleep gamma power across the 24h period since these sleep states were not represented in every hour. Student s t-tests were used comparing o.d. for SAL- and VPA-exposed tissue. One-way ANOVA followed by a Newman-Keuls post-hoc test was used to compare the o.d. values for SAL- and VPA-exposed GAD expression by tail malformation severity. 115

131 Results VPA exposure increases wakefulness at the expense of sleep in juvenile rats. Nocturnal adult rodents, like the Spague-Dawley rat, cycle through bouts of sleep (NREMS and REMS) and wake in both the dark (active) and light (quiescent) phases. Typically, a higher percentage of accumulated sleep occurs in the light phase with consolidated bouts of wake occurring in the dark phase. Visual inspection of hypnograms generated from the scored EEG traces from SAL- and VPA-exposed animals suggested the juvenile controls (SAL-exposed) followed a normal adult-like pattern with consolidated bouts of wake primarily present in the dark phase. In contrast, the VPAexposed rats appeared to experience consolidated bouts of wake in both the light and dark phases (Fig. 31). Quantitative analysis of the scored traces for the total time spent in NREMS, REMS and wake revealed significant differences between the SAL- and VPA-exposed groups. Across the 12h of the light phase, the VPA-exposed animals spent significantly more time in wake (~35 minutes; t 9 = 2.741, P = 0.023; Fig. 32A) with no significant differences in the dark phase (t 9 = 0.788, P > 0.05). Moreover, across the 24h light:dark period, VPA-exposed rats again spent significantly more time in wake with the difference increasing to approximately 48 minutes compared to the SAL-exposed controls (t 9 = 2.382, P = 0.041; Fig. 32A). As a consequence of the increased wake time in the light phase, total sleep time (TST; sum of the time spent in NREMS and REMS) decreased significantly in the VPAexposed animals by approximately 38 minutes compared to controls (t 9 = 3.126, P = 0.012; data not shown). Changes in NREMS accounted for the overall difference in TST. 116

132 Figure 31. Representative hypnograms of juvenile rats exposed to SAL and VPA in utero across a 24h period. W = wake, N = NREMS, and R = REMS. The solid black bar is highlighting consolidated wake bouts during the light phase of VPA-exposed rats, which are absent in the SAL-controls. 117

133 In the light phase, the VPA-exposed animals spent significantly less time in NREMS (~30 minutes; t 9 = 2.621, P = 0.028; Fig. 32B) with no significant differences in the dark phase (t 9 = 0.553, P = 0.594), whereas REMS was unaffected by VPA-exposure in either the light or dark phase (Fig. 2C; t 9 = 0.920, P = 0.382, light phase and t 9 = 1.905, P = 0.09, dark phase). In contrast to individual 12hr phases, the 24h total of REMS revealed that VPA-exposed animals spent significant less time in REMS (~17 minutes) compared to the controls (t 9 = 2.799, P = 0.021; Fig. 32C). VPA exposure disrupts the pattern of juvenile sleep/wake cycles. In utero exposure to VPA markedly changed the sleep/wake architecture (mean bout duration, bout number and number of transitions into Wake, NREMS or REMS) compared to controls. VPA exposure significantly increased the mean duration of a wake bout in the light and dark phases by ~117% and ~70%, respectively, compared to the SAL-exposed animals (t 9 = 6.966, P < , light phase and t 9 = 2.552, P < 0.031, dark phase; Fig. 33A). The increase in bout duration was associated with a significant decrease in the number of wake bouts in the light phase only (t 9 = 6.532, P = , Fig. 33A). For sleep, only NREMS demonstrated significant changes in architecture. Curiously, in the light phase, VPA exposure significantly increased the average duration of a NREMS bout by ~35% (t 9 = 3.517, P = 0.007; Fig. 33B) while the number of NREMS bouts was significantly reduced by ~32% compared to SAL-exposed animals (t 9 = 6.646, P < , Fig. 33B). For REMS, there were no significant differences in the mean bout duration or bout number in either phase (data not shown). 118

134 Figure 32. Effects of in utero exposure to VPA on sleep-wake behavior in juvenile rats. (A) During the light phase, VPA-exposed rats spend more time in wake (~35 minutes) compared to SAL-exposed rats. Across the 24h period, VPA-exposed rats spend ~48 minutes more in wake than SAL-exposed controls. (B) VPA exposure reduced NREMS time by ~30 minutes during the light phase only. (C) REMS during the individual light and dark phases was not affected by VPAexposure, however; the 24hr REMS total is reduced by ~ 17 minutes in the VPA-exposed juveniles compared to SAL-exposed controls. *, P < 0.05 vs. SAL-exposed controls. Data are represented as mean ± SEM. 119

135 Figure 33. Effects of in utero exposure to VPA on sleep-wake architecture in juvenile rats. (Atop) During both the light and dark phases, VPA-exposed rats have, on average, longer bouts of wakefulness. (A-bottom) The increased duration of wake bouts is associated with a decrease in the number of wake bouts during the light phase only. (B-top) NREMS bouts, on average, are significantly increased in VPA-exposed rats as well during the light phase. (B-bottom) This increase in NREMS bout duration is also associated with a decrease in the number of NREMS bouts during the light phase. *, P < 0.05 vs. SAL-exposed controls. Data are represented as mean ± SEM. 120

136 To assess whether VPA exposure affects the stability of maintaining a wake or sleep state, the number of transitions into and out of wake, NREMS and REMS were quantified across the light and dark phases. In the light phase, the number of transitions to and from wake was significantly reduced by approximately 50% in VPA exposed animals compared to the SAL exposed animals (Table 5; wake to NREMS: t 9 = 6.460, P = , NREMS to wake: t 9 = 6.509, P = , and REM to wake: t 9 = 3.484, P = 0.007). No significant differences were observed in the dark phase or when transitioning between sleep states (Table 5). VPA-exposure results in changes to the power spectra. Power spectral analysis was utilized to study the cortical EEG of the juvenile rats exposed to VPA or SAL in utero during the light and dark phase. We also investigated differences in power in wake, NREMS, and REMS. In the spectral analysis of all frequencies in the continuous EEG, VPA-exposure reduced the theta power in both the 12hr light and dark phases (Table 6). When we calculated the mean power for each frequency band in for the individual vigilance states (wake, NREMS, and REMS) we Table 5. Effects of in utero exposure to VPA on vigilance state transitions. Light Phase Dark Phase Transition SAL VPA SAL VPA Wake to NREMS ± ± 5.3* 55.5 ± ± 6.5 NREMS to Wake 83.2 ± ± 4.5* 39.5 ± ± 5.1 NREMS to REM 66.8 ± ± ± ± 3.7 REMS to Wake 20.5 ± ± 1.1* 17.2 ± ± 1.7 REMS to NREMS 46.2 ± ± ± ± 4.3 Data are represented as mean ± SEM. *, P < 0.05 vs. SAL 121

137 found that VPA exposure reduced mid-range frequencies while increasing higher range frequencies primarily in wake and REMS; the two states characterized by EEG desynchrony (Fig. 34A & C and Table 7). In all vigilance states, VPA exposure decreased theta band densities (4-8hz) in both the light phase (wake: U = 0.0, P = 0.004; NREMS: U = 2.0, P = 0.017; REMS: U=0.0, P = 0.004; Fig. 34A-C) and the dark phase (wake: U = 2.0, P = 0.017; NREMS: U = 3.0, P = 0.03; REMS: U = 0.0, P = 0.004); Fig. 34A-C). Most notable was the approximate 20% decrease in theta band density during REMS regardless of phase (Fig. 34C). Alpha band densities (8-12hz) were also significantly decreased during wake in the light (9.5%) and dark (13.%) phases in VPAexposed rats compared to SAL-exposed controls (Table 7; light: U = 1.0, P = 0.009; dark: U = 0.0, P = 0.004). Conversely, VPA exposure increased high frequency gamma band densities (30-80 Hz) during wake and REMS in the light phase (Figure 34A & C; wake: U = 1.0, P = 0.009; REMS: U = 0.0, P = 0.004) and dark phase (wake: U = 0.0, P = 0.004; REMS: U = 0.0, P = 0.004) compared to the SAL-exposed controls. Interestingly, the greatest magnitude of change in the frequency spectrum occurred in REMS gamma power in the light (~50%) and dark (~37%) phases. To a lesser degree, sigma and beta, also increased during wake and sleep, primarily REMS, in the light phase (Table 7; wake: sigma, U = 0.0, P = 0.004; beta, U = 4.0, P = 0.052, and REMS: sigma, U = 4.0, P = 0.052; beta: U = 3.0, P = 0.03) and dark phase (Table 7; wake: sigma, U = 3.0, P = 0.03; beta, U = 0.0, P = 0.004; NREMS: sigma. U = 4.0, P = 0.052; and REMS: beta, U = 3.0, P = 0.004) in VPA-exposed animals. Again, the most prominent changes were during REMS; there was approximately a 30% change in beta and a 22% change in sigma. 122

138 Figure 34. Effects of in utero exposure to VPA on cortical power spectra in juvenile rats. The average theta power in (A) wake, (B) NREMS, and (C) REMS is reduced during both the light and dark phases of VPA-exposed rats. In both the light and dark phases, gamma power is increased during (A) wake and (C) REMS in VPA-exposed rats compared to SAL-exposed controls. *, P < 0.05 vs. SAL-exposed controls. Data are represented as mean ± SEM. 123

139 Table 6. Effects of in utero exposure to VPA on averaged power spectra (% total power) in the light and dark phase. Power Band (Hz) Light Phase Dark Phase SAL VPA SAL VPA Delta (0.5-4) 44.9 ± ± ± ± 1.7 Theta (4-8) ± ± 0.96* 31.0 ± ± 0.7* Alpha (8-12) 10.5 ± ± ± ± 0.7 Sigma (12-16) 4.2 ± ± ± ± 0.6 Beta (16-24) 4.4 ± ± ± ± 0.4 Gamma (30-80) 7.2 ± ± ± ± 0.3 Data are represented as mean ± SEM. *, P < 0.05 vs. SAL Since spectral analysis removes time as a factor, it is difficult to address whether changes in the mean spectral powers are due to the power itself or longer/shorter durations of the vigilance states. To address this, we specifically analyzed the mean gamma power by hour across the 24h period. There was a main effect (F 1,23 = 14.50; P = 0.004; Fig. 34D) of in utero exposure on wake gamma power across the 24h period. Similarly, there was a main effect (F 1,23 = 147.9; P < ; Fig. 34F) of in utero exposure on REMS gamma power across the 24h period. VPA-exposed animals had elevated gamma power across the 24h period during wake and REMS bouts compared to SAL-exposed animals. There was no significant difference in NREM sleep gamma power across the 24h period (Fig. 33E). VPA-exposure results in changes to the cortical GABAergic system. In the present study, we observed varying degrees of tail malformations in the VPA-exposed animals that included barely discernable changes (Fig. 35A; arrows) or 124

140 Table 7. Effects of in utero exposure to VPA on mean power (% total power) for alpha, sigma, and beta frequency bands Power Band (Hz) WAKE Light Phase Dark Phase SAL VPA SAL VPA Alpha (8-12) 6.3 ± ± 0.1* 6.9 ± ± 0.1* Sigma (12-16) 2.6 ± ± 0.05* 2.7 ± ± 0.05* Beta (16-24) 3.5 ± ± ± ± 0.1* NREMS Light Phase Dark Phase SAL VPA SAL VPA Alpha (8-12) 5.9 ± ± ± ± 0.3 Sigma (12-16) 2.4 ± ± ± ± 0.3* Beta (16-24) 2.1 ± ± ± ± 0.2 REMS Light Phase Dark Phase SAL VPA SAL VPA Alpha (8-12) 7.8 ± ± ± ± 0.3 Sigma (12-16) 3.2 ± ± 0.2* 3.2 ± ± 0.3 Beta (16-24) 4.0 ± ± 0.3* 4.0 ± ± 0.2* Data are represented as mean ± SEM. *, P < 0.05 vs. SAL kinks and bends (Fig 5A; asterisks). Western blot analysis of protein isolated from primary somatosensory cortex micropunches from all animals revealed no differences in the cortical GABAergic system in VPA-treated animals compared to SAL-treated controls (GAD67: t 11 = , P > 0.05; GAD65: t 11 = ; P > 0.05; Fig. 35B & C). Interestingly, the o.d. values for the two VPA-exposed animals with the barely discernable tail malformations (n=2, circled in Fig. 35C) appeared more similar to the controls. Thus, if the VPA-exposed animals were re-grouped according to the severity of the tail malformations (minor, M and severe, S), then both isoforms of GAD (65 and 67) were significantly reduced in VPA-exposed animals with severe tail malformations (GAD67: F 2,10 = 6.294, P = 0.02; GAD65: F 2,10 = 4.616, P = 0.04) compared to SAL 125

141 Figure 35. Effects of in utero exposure to VPA on cortical expression of GAD65 and GAD67 in juvenile rats. (A) Images of the tails of juvenile rats exposed to either SAL or VPA. The tails of VPA-exposed rats had varying degrees of malformations, including barely discernable changes (arrows; referred to as minor) to more severe kinks and bends (asterisks; referred to as severe). (B) Representative immunoblots from micropunches of the primary somatosensory cortex of SAL- and VPA-exposed rats. (C) Plots of the individual o.d. values from Sal and VPA exposed animals. The circled points are values from VPA-exposed animals with minor tail malformations. (D) GAD65 and GAD67 protein expression is reduced in VPA-exposed rats with severe tail malformations compared to SAL-exposed controls and VPA-exposed animals with minor tail malformations. *, P < 0.05 vs. SAL-exposed controls and VPA-exposed animals with minor tail malformations. Data are represented as mean ± SEM. 126